1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-27 03:21:53 +03:00
samba-mirror/source4/rpc_server/drsuapi/getncchanges.c

3608 lines
107 KiB
C
Raw Normal View History

/*
Unix SMB/CIFS implementation.
implement the DSGetNCChanges call
Copyright (C) Anatoliy Atanasov 2009
Copyright (C) Andrew Tridgell 2009-2010
Copyright (C) Andrew Bartlett 2010-2016
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "includes.h"
#include "rpc_server/dcerpc_server.h"
#include "dsdb/samdb/samdb.h"
#include "param/param.h"
#include "librpc/gen_ndr/ndr_drsblobs.h"
#include "librpc/gen_ndr/ndr_drsuapi.h"
#include "librpc/gen_ndr/ndr_security.h"
#include "libcli/security/security.h"
#include "libcli/security/session.h"
#include "rpc_server/drsuapi/dcesrv_drsuapi.h"
#include "rpc_server/common/sid_helper.h"
#include "../libcli/drsuapi/drsuapi.h"
#include "lib/util/binsearch.h"
#include "lib/util/tsort.h"
#include "auth/session.h"
#include "dsdb/common/util.h"
#include "lib/dbwrap/dbwrap.h"
#include "lib/dbwrap/dbwrap_rbt.h"
#include "librpc/gen_ndr/ndr_misc.h"
#undef DBGC_CLASS
#define DBGC_CLASS DBGC_DRS_REPL
#define DRS_GUID_SIZE 16
#define DEFAULT_MAX_OBJECTS 1000
#define DEFAULT_MAX_LINKS 1500
/*
* state of a partially-completed replication cycle. This state persists
* over multiple calls to dcesrv_drsuapi_DsGetNCChanges()
*/
struct drsuapi_getncchanges_state {
struct db_context *obj_cache;
struct GUID *guids;
uint32_t num_records;
uint32_t num_processed;
struct ldb_dn *ncRoot_dn;
struct GUID ncRoot_guid;
bool is_schema_nc;
bool is_get_anc;
bool is_get_tgt;
uint64_t min_usn;
uint64_t max_usn;
struct drsuapi_DsReplicaHighWaterMark last_hwm;
struct ldb_dn *last_dn;
struct drsuapi_DsReplicaHighWaterMark final_hwm;
struct drsuapi_DsReplicaCursor2CtrEx *final_udv;
struct drsuapi_DsReplicaLinkedAttribute *la_list;
uint32_t la_count;
uint32_t la_idx;
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
/* these are just used for debugging the replication's progress */
uint32_t links_given;
uint32_t total_links;
};
/* We must keep the GUIDs in NDR form for sorting */
struct la_for_sorting {
const struct drsuapi_DsReplicaLinkedAttribute *link;
uint8_t target_guid[DRS_GUID_SIZE];
uint8_t source_guid[DRS_GUID_SIZE];
};
/*
* stores the state for a chunk of replication data. This state information
* only exists for a single call to dcesrv_drsuapi_DsGetNCChanges()
*/
struct getncchanges_repl_chunk {
uint32_t max_objects;
uint32_t max_links;
uint32_t tgt_la_count;
bool immediate_link_sync;
time_t max_wait;
time_t start;
/* stores the objects to be sent in this chunk */
uint32_t object_count;
struct drsuapi_DsReplicaObjectListItemEx *object_list;
/* the last object added to this replication chunk */
struct drsuapi_DsReplicaObjectListItemEx *last_object;
};
static int drsuapi_DsReplicaHighWaterMark_cmp(const struct drsuapi_DsReplicaHighWaterMark *h1,
const struct drsuapi_DsReplicaHighWaterMark *h2)
{
if (h1->highest_usn < h2->highest_usn) {
return -1;
} else if (h1->highest_usn > h2->highest_usn) {
return 1;
} else if (h1->tmp_highest_usn < h2->tmp_highest_usn) {
return -1;
} else if (h1->tmp_highest_usn > h2->tmp_highest_usn) {
return 1;
} else if (h1->reserved_usn < h2->reserved_usn) {
return -1;
} else if (h1->reserved_usn > h2->reserved_usn) {
return 1;
}
return 0;
}
/*
build a DsReplicaObjectIdentifier from a ldb msg
*/
static struct drsuapi_DsReplicaObjectIdentifier *get_object_identifier(TALLOC_CTX *mem_ctx,
const struct ldb_message *msg)
{
struct drsuapi_DsReplicaObjectIdentifier *identifier;
struct dom_sid *sid;
identifier = talloc(mem_ctx, struct drsuapi_DsReplicaObjectIdentifier);
if (identifier == NULL) {
return NULL;
}
identifier->dn = ldb_dn_alloc_linearized(identifier, msg->dn);
identifier->guid = samdb_result_guid(msg, "objectGUID");
sid = samdb_result_dom_sid(identifier, msg, "objectSid");
if (sid) {
identifier->sid = *sid;
} else {
ZERO_STRUCT(identifier->sid);
}
return identifier;
}
static int udv_compare(const struct GUID *guid1, struct GUID guid2)
{
return GUID_compare(guid1, &guid2);
}
/*
see if we can filter an attribute using the uptodateness_vector
*/
static bool udv_filter(const struct drsuapi_DsReplicaCursorCtrEx *udv,
const struct GUID *originating_invocation_id,
uint64_t originating_usn)
{
const struct drsuapi_DsReplicaCursor *c;
if (udv == NULL) return false;
BINARY_ARRAY_SEARCH(udv->cursors, udv->count, source_dsa_invocation_id,
originating_invocation_id, udv_compare, c);
if (c && originating_usn <= c->highest_usn) {
return true;
}
return false;
}
static int uint32_t_cmp(uint32_t a1, uint32_t a2)
{
if (a1 == a2) return 0;
return a1 > a2 ? 1 : -1;
}
static int uint32_t_ptr_cmp(uint32_t *a1, uint32_t *a2)
{
if (*a1 == *a2) return 0;
return *a1 > *a2 ? 1 : -1;
}
static WERROR getncchanges_attid_remote_to_local(const struct dsdb_schema *schema,
const struct dsdb_syntax_ctx *ctx,
enum drsuapi_DsAttributeId remote_attid_as_enum,
enum drsuapi_DsAttributeId *local_attid_as_enum,
const struct dsdb_attribute **_sa)
{
WERROR werr;
const struct dsdb_attribute *sa = NULL;
if (ctx->pfm_remote == NULL) {
DEBUG(7, ("No prefixMap supplied, falling back to local prefixMap.\n"));
goto fail;
}
werr = dsdb_attribute_drsuapi_remote_to_local(ctx,
remote_attid_as_enum,
local_attid_as_enum,
_sa);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(3, ("WARNING: Unable to resolve remote attid, falling back to local prefixMap.\n"));
goto fail;
}
return werr;
fail:
sa = dsdb_attribute_by_attributeID_id(schema, remote_attid_as_enum);
if (sa == NULL) {
return WERR_DS_DRA_SCHEMA_MISMATCH;
} else {
if (local_attid_as_enum != NULL) {
*local_attid_as_enum = sa->attributeID_id;
}
if (_sa != NULL) {
*_sa = sa;
}
return WERR_OK;
}
}
static WERROR getncchanges_update_revealed_list(struct ldb_context *sam_ctx,
TALLOC_CTX *mem_ctx,
struct ldb_message **msg,
struct ldb_dn *object_dn,
const struct GUID *object_guid,
const struct dsdb_attribute *sa,
struct replPropertyMetaData1 *meta_data,
struct ldb_message *revealed_users)
{
enum ndr_err_code ndr_err;
int ldb_err;
char *attr_str = NULL;
char *attr_hex = NULL;
DATA_BLOB attr_blob;
struct ldb_message_element *existing = NULL, *el_add = NULL, *el_del = NULL;
const char * const * secret_attributes = ldb_get_opaque(sam_ctx, "LDB_SECRET_ATTRIBUTE_LIST");
if (!ldb_attr_in_list(secret_attributes,
sa->lDAPDisplayName)) {
return WERR_OK;
}
ndr_err = ndr_push_struct_blob(&attr_blob, mem_ctx, meta_data, (ndr_push_flags_fn_t)ndr_push_replPropertyMetaData1);
if (!NDR_ERR_CODE_IS_SUCCESS(ndr_err)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
attr_hex = hex_encode_talloc(mem_ctx, attr_blob.data, attr_blob.length);
if (attr_hex == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
attr_str = talloc_asprintf(mem_ctx, "B:%zd:%s:%s", attr_blob.length*2, attr_hex, ldb_dn_get_linearized(object_dn));
if (attr_str == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
existing = ldb_msg_find_element(revealed_users, "msDS-RevealedUsers");
if (existing != NULL) {
/* Replace the old value (if one exists) with the current one */
struct parsed_dn *link_dns;
struct parsed_dn *exact = NULL, *unused = NULL;
uint8_t attid[4];
DATA_BLOB partial_meta;
ldb_err = get_parsed_dns_trusted(mem_ctx, existing, &link_dns);
if (ldb_err != LDB_SUCCESS) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
/* Construct a partial metadata blob to match on in the DB */
SIVAL(attid, 0, sa->attributeID_id);
partial_meta.length = 4;
partial_meta.data = attid;
/* Binary search using GUID and attribute id for uniqueness */
ldb_err = parsed_dn_find(sam_ctx, link_dns, existing->num_values,
object_guid, object_dn,
partial_meta, 4,
&exact, &unused,
DSDB_SYNTAX_BINARY_DN, true);
if (ldb_err != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed parsed DN find - %s\n",
ldb_errstring(sam_ctx)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (exact != NULL) {
/* Perform some verification of the blob */
struct replPropertyMetaData1 existing_meta_data;
ndr_err = ndr_pull_struct_blob_all_noalloc(&exact->dsdb_dn->extra_part,
&existing_meta_data,
(ndr_pull_flags_fn_t)ndr_pull_replPropertyMetaData1);
if (!NDR_ERR_CODE_IS_SUCCESS(ndr_err)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (existing_meta_data.attid == sa->attributeID_id) {
ldb_err = ldb_msg_add_empty(*msg, "msDS-RevealedUsers", LDB_FLAG_MOD_DELETE, &el_del);
if (ldb_err != LDB_SUCCESS) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
el_del->values = talloc_array((*msg)->elements, struct ldb_val, 1);
if (el_del->values == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
el_del->values[0] = *exact->v;
el_del->num_values = 1;
} else {
return WERR_DS_DRA_INTERNAL_ERROR;
}
}
}
ldb_err = ldb_msg_add_empty(*msg, "msDS-RevealedUsers", LDB_FLAG_MOD_ADD, &el_add);
if (ldb_err != LDB_SUCCESS) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
el_add->values = talloc_array((*msg)->elements, struct ldb_val, 1);
if (el_add->values == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
el_add->values[0] = data_blob_string_const(attr_str);
el_add->num_values = 1;
return WERR_OK;
}
/*
* This function filter attributes for build_object based on the
* uptodatenessvector and partial attribute set.
*
* Any secret attributes are forced here for REPL_SECRET, and audited at this
* point with msDS-RevealedUsers.
*/
static WERROR get_nc_changes_filter_attrs(struct drsuapi_DsReplicaObjectListItemEx *obj,
struct replPropertyMetaDataBlob md,
struct ldb_context *sam_ctx,
const struct ldb_message *msg,
const struct GUID *guid,
uint32_t *count,
uint64_t highest_usn,
const struct dsdb_attribute *rdn_sa,
struct dsdb_schema *schema,
struct drsuapi_DsReplicaCursorCtrEx *uptodateness_vector,
struct drsuapi_DsPartialAttributeSet *partial_attribute_set,
uint32_t *local_pas,
uint32_t *attids,
bool exop_secret,
struct ldb_message **revealed_list_msg,
struct ldb_message *existing_revealed_list_msg)
{
uint32_t i, n;
WERROR werr;
for (n=i=0; i<md.ctr.ctr1.count; i++) {
const struct dsdb_attribute *sa;
bool force_attribute = false;
/* if the attribute has not changed, and it is not the
instanceType then don't include it */
if (md.ctr.ctr1.array[i].local_usn < highest_usn &&
!exop_secret &&
md.ctr.ctr1.array[i].attid != DRSUAPI_ATTID_instanceType) continue;
/* don't include the rDN */
if (md.ctr.ctr1.array[i].attid == rdn_sa->attributeID_id) continue;
sa = dsdb_attribute_by_attributeID_id(schema, md.ctr.ctr1.array[i].attid);
if (!sa) {
DEBUG(0,(__location__ ": Failed to find attribute in schema for attrid %u mentioned in replPropertyMetaData of %s\n",
(unsigned int)md.ctr.ctr1.array[i].attid,
ldb_dn_get_linearized(msg->dn)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (sa->linkID) {
struct ldb_message_element *el;
el = ldb_msg_find_element(msg, sa->lDAPDisplayName);
if (el && el->num_values && dsdb_dn_is_upgraded_link_val(&el->values[0])) {
/* don't send upgraded links inline */
continue;
}
}
if (exop_secret &&
!dsdb_attr_in_rodc_fas(sa)) {
force_attribute = true;
DEBUG(4,("Forcing attribute %s in %s\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
werr = getncchanges_update_revealed_list(sam_ctx, obj,
revealed_list_msg,
msg->dn, guid, sa,
&md.ctr.ctr1.array[i],
existing_revealed_list_msg);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
/* filter by uptodateness_vector */
if (md.ctr.ctr1.array[i].attid != DRSUAPI_ATTID_instanceType &&
!force_attribute &&
udv_filter(uptodateness_vector,
&md.ctr.ctr1.array[i].originating_invocation_id,
md.ctr.ctr1.array[i].originating_usn)) {
continue;
}
/* filter by partial_attribute_set */
if (partial_attribute_set && !force_attribute) {
uint32_t *result = NULL;
BINARY_ARRAY_SEARCH_V(local_pas, partial_attribute_set->num_attids, sa->attributeID_id,
uint32_t_cmp, result);
if (result == NULL) {
continue;
}
}
obj->meta_data_ctr->meta_data[n].originating_change_time = md.ctr.ctr1.array[i].originating_change_time;
obj->meta_data_ctr->meta_data[n].version = md.ctr.ctr1.array[i].version;
obj->meta_data_ctr->meta_data[n].originating_invocation_id = md.ctr.ctr1.array[i].originating_invocation_id;
obj->meta_data_ctr->meta_data[n].originating_usn = md.ctr.ctr1.array[i].originating_usn;
attids[n] = md.ctr.ctr1.array[i].attid;
n++;
}
*count = n;
return WERR_OK;
}
/*
drsuapi_DsGetNCChanges for one object
*/
static WERROR get_nc_changes_build_object(struct drsuapi_DsReplicaObjectListItemEx *obj,
const struct ldb_message *msg,
struct ldb_context *sam_ctx,
struct drsuapi_getncchanges_state *getnc_state,
struct dsdb_schema *schema,
DATA_BLOB *session_key,
struct drsuapi_DsGetNCChangesRequest10 *req10,
bool force_object_return,
uint32_t *local_pas,
struct ldb_dn *machine_dn,
const struct GUID *guid)
{
const struct ldb_val *md_value;
uint32_t i, n;
struct replPropertyMetaDataBlob md;
uint32_t rid = 0;
int ldb_err;
enum ndr_err_code ndr_err;
uint32_t *attids;
const char *rdn;
const struct dsdb_attribute *rdn_sa;
uint64_t uSNChanged;
unsigned int instanceType;
struct dsdb_syntax_ctx syntax_ctx;
struct ldb_result *res = NULL;
WERROR werr;
int ret;
uint32_t replica_flags = req10->replica_flags;
struct drsuapi_DsPartialAttributeSet *partial_attribute_set =
req10->partial_attribute_set;
struct drsuapi_DsReplicaCursorCtrEx *uptodateness_vector =
req10->uptodateness_vector;
enum drsuapi_DsExtendedOperation extended_op = req10->extended_op;
bool is_schema_nc = getnc_state->is_schema_nc;
uint64_t highest_usn = getnc_state->min_usn;
/* make dsdb sytanx context for conversions */
dsdb_syntax_ctx_init(&syntax_ctx, sam_ctx, schema);
syntax_ctx.is_schema_nc = is_schema_nc;
uSNChanged = ldb_msg_find_attr_as_uint64(msg, "uSNChanged", 0);
instanceType = ldb_msg_find_attr_as_uint(msg, "instanceType", 0);
if (instanceType & INSTANCE_TYPE_IS_NC_HEAD) {
obj->is_nc_prefix = true;
obj->parent_object_guid = NULL;
} else {
obj->is_nc_prefix = false;
obj->parent_object_guid = talloc(obj, struct GUID);
if (obj->parent_object_guid == NULL) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
*obj->parent_object_guid = samdb_result_guid(msg, "parentGUID");
if (GUID_all_zero(obj->parent_object_guid)) {
DEBUG(0,(__location__ ": missing parentGUID for %s\n",
ldb_dn_get_linearized(msg->dn)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
}
obj->next_object = NULL;
md_value = ldb_msg_find_ldb_val(msg, "replPropertyMetaData");
if (!md_value) {
/* nothing to send */
return WERR_OK;
}
if (instanceType & INSTANCE_TYPE_UNINSTANT) {
/* don't send uninstantiated objects */
return WERR_OK;
}
if (uSNChanged <= highest_usn) {
/* nothing to send */
return WERR_OK;
}
ndr_err = ndr_pull_struct_blob(md_value, obj, &md,
(ndr_pull_flags_fn_t)ndr_pull_replPropertyMetaDataBlob);
if (!NDR_ERR_CODE_IS_SUCCESS(ndr_err)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (md.version != 1) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
rdn = ldb_dn_get_rdn_name(msg->dn);
if (rdn == NULL) {
DEBUG(0,(__location__ ": No rDN for %s\n", ldb_dn_get_linearized(msg->dn)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
rdn_sa = dsdb_attribute_by_lDAPDisplayName(schema, rdn);
if (rdn_sa == NULL) {
DEBUG(0,(__location__ ": Can't find dsds_attribute for rDN %s in %s\n",
rdn, ldb_dn_get_linearized(msg->dn)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
obj->meta_data_ctr = talloc(obj, struct drsuapi_DsReplicaMetaDataCtr);
attids = talloc_array(obj, uint32_t, md.ctr.ctr1.count);
obj->object.identifier = get_object_identifier(obj, msg);
if (obj->object.identifier == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
dom_sid_split_rid(NULL, &obj->object.identifier->sid, NULL, &rid);
obj->meta_data_ctr->meta_data = talloc_array(obj, struct drsuapi_DsReplicaMetaData, md.ctr.ctr1.count);
if (extended_op == DRSUAPI_EXOP_REPL_SECRET) {
/* Get the existing revealed users for the destination */
struct ldb_message *revealed_list_msg = NULL;
struct ldb_message *existing_revealed_list_msg = NULL;
const char *machine_attrs[] = {
"msDS-RevealedUsers",
NULL
};
revealed_list_msg = ldb_msg_new(sam_ctx);
if (revealed_list_msg == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
revealed_list_msg->dn = machine_dn;
ret = ldb_transaction_start(sam_ctx);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction start - %s\n",
ldb_errstring(sam_ctx)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
ldb_err = dsdb_search_dn(sam_ctx, obj, &res, machine_dn, machine_attrs, DSDB_SEARCH_SHOW_EXTENDED_DN);
if (ldb_err != LDB_SUCCESS || res->count != 1) {
ldb_transaction_cancel(sam_ctx);
return WERR_DS_DRA_INTERNAL_ERROR;
}
existing_revealed_list_msg = res->msgs[0];
werr = get_nc_changes_filter_attrs(obj, md, sam_ctx, msg,
guid, &n, highest_usn,
rdn_sa, schema,
uptodateness_vector,
partial_attribute_set, local_pas,
attids,
true,
&revealed_list_msg,
existing_revealed_list_msg);
if (!W_ERROR_IS_OK(werr)) {
ldb_transaction_cancel(sam_ctx);
return werr;
}
if (revealed_list_msg != NULL) {
ret = ldb_modify(sam_ctx, revealed_list_msg);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed to alter revealed links - %s\n",
ldb_errstring(sam_ctx)));
ldb_transaction_cancel(sam_ctx);
return WERR_DS_DRA_INTERNAL_ERROR;
}
}
ret = ldb_transaction_commit(sam_ctx);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction commit - %s\n",
ldb_errstring(sam_ctx)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
} else {
werr = get_nc_changes_filter_attrs(obj, md, sam_ctx, msg, guid,
&n, highest_usn, rdn_sa,
schema, uptodateness_vector,
partial_attribute_set, local_pas,
attids,
false,
NULL,
NULL);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
/* ignore it if its an empty change. Note that renames always
* change the 'name' attribute, so they won't be ignored by
* this
* the force_object_return check is used to force an empty
* object return when we timeout in the getncchanges loop.
* This allows us to return an empty object, which keeps the
* client happy while preventing timeouts
*/
if (n == 0 ||
(n == 1 &&
attids[0] == DRSUAPI_ATTID_instanceType &&
!force_object_return)) {
talloc_free(obj->meta_data_ctr);
obj->meta_data_ctr = NULL;
return WERR_OK;
}
obj->meta_data_ctr->count = n;
obj->object.flags = DRSUAPI_DS_REPLICA_OBJECT_FROM_MASTER;
obj->object.attribute_ctr.num_attributes = obj->meta_data_ctr->count;
obj->object.attribute_ctr.attributes = talloc_array(obj, struct drsuapi_DsReplicaAttribute,
obj->object.attribute_ctr.num_attributes);
if (obj->object.attribute_ctr.attributes == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
/*
* Note that the meta_data array and the attributes array must
* be the same size and in the same order
*/
for (i=0; i<obj->object.attribute_ctr.num_attributes; i++) {
struct ldb_message_element *el;
const struct dsdb_attribute *sa;
sa = dsdb_attribute_by_attributeID_id(schema, attids[i]);
if (!sa) {
DEBUG(0,("Unable to find attributeID %u in schema\n", attids[i]));
return WERR_DS_DRA_INTERNAL_ERROR;
}
el = ldb_msg_find_element(msg, sa->lDAPDisplayName);
if (el == NULL) {
/* this happens for attributes that have been removed */
DEBUG(5,("No element '%s' for attributeID %u in message\n",
sa->lDAPDisplayName, attids[i]));
ZERO_STRUCT(obj->object.attribute_ctr.attributes[i]);
obj->object.attribute_ctr.attributes[i].attid =
dsdb_attribute_get_attid(sa, syntax_ctx.is_schema_nc);
} else {
werr = sa->syntax->ldb_to_drsuapi(&syntax_ctx, sa, el, obj,
&obj->object.attribute_ctr.attributes[i]);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,("Unable to convert %s on %s to DRS object - %s\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn),
win_errstr(werr)));
return werr;
}
/* if DRSUAPI_DRS_SPECIAL_SECRET_PROCESSING is set
* check if attribute is secret and send a null value
*/
if (replica_flags & DRSUAPI_DRS_SPECIAL_SECRET_PROCESSING) {
drsuapi_process_secret_attribute(&obj->object.attribute_ctr.attributes[i],
&obj->meta_data_ctr->meta_data[i]);
}
/* some attributes needs to be encrypted
before being sent */
werr = drsuapi_encrypt_attribute(obj, session_key, rid,
&obj->object.attribute_ctr.attributes[i]);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,("Unable to encrypt %s on %s in DRS object - %s\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn),
win_errstr(werr)));
return werr;
}
}
if (attids[i] != obj->object.attribute_ctr.attributes[i].attid) {
DEBUG(0, ("Unable to replicate attribute %s on %s via DRS, incorrect attributeID: "
"0x%08x vs 0x%08x "
"Run dbcheck!\n",
sa->lDAPDisplayName,
ldb_dn_get_linearized(msg->dn),
attids[i],
obj->object.attribute_ctr.attributes[i].attid));
return WERR_DS_DATABASE_ERROR;
}
}
return WERR_OK;
}
/*
add one linked attribute from an object to the list of linked
attributes in a getncchanges request
*/
static WERROR get_nc_changes_add_la(TALLOC_CTX *mem_ctx,
struct ldb_context *sam_ctx,
const struct dsdb_schema *schema,
const struct dsdb_attribute *sa,
const struct ldb_message *msg,
struct dsdb_dn *dsdb_dn,
struct drsuapi_DsReplicaLinkedAttribute **la_list,
uint32_t *la_count,
bool is_schema_nc)
{
struct drsuapi_DsReplicaLinkedAttribute *la;
bool active;
NTSTATUS status;
WERROR werr;
(*la_list) = talloc_realloc(mem_ctx, *la_list, struct drsuapi_DsReplicaLinkedAttribute, (*la_count)+1);
W_ERROR_HAVE_NO_MEMORY(*la_list);
la = &(*la_list)[*la_count];
la->identifier = get_object_identifier(*la_list, msg);
W_ERROR_HAVE_NO_MEMORY(la->identifier);
active = (dsdb_dn_rmd_flags(dsdb_dn->dn) & DSDB_RMD_FLAG_DELETED) == 0;
if (!active) {
/* We have to check that the inactive link still point to an existing object */
struct GUID guid;
struct ldb_dn *tdn;
int ret;
const char *v;
v = ldb_msg_find_attr_as_string(msg, "isDeleted", "FALSE");
if (strncmp(v, "TRUE", 4) == 0) {
/*
* Note: we skip the transmition of the deleted link even if the other part used to
* know about it because when we transmit the deletion of the object, the link will
* be deleted too due to deletion of object where link points and Windows do so.
*/
if (dsdb_functional_level(sam_ctx) >= DS_DOMAIN_FUNCTION_2008_R2) {
v = ldb_msg_find_attr_as_string(msg, "isRecycled", "FALSE");
/*
* On Windows 2008R2 isRecycled is always present even if FL or DL are < FL 2K8R2
* if it join an existing domain with deleted objets, it firsts impose to have a
* schema with the is-Recycled object and for all deleted objects it adds the isRecycled
* either during initial replication or after the getNCChanges.
* Behavior of samba has been changed to always have this attribute if it's present in the schema.
*
* So if FL <2K8R2 isRecycled might be here or not but we don't care, it's meaning less.
* If FL >=2K8R2 we are sure that this attribute will be here.
* For this kind of forest level we do not return the link if the object is recycled
* (isRecycled = true).
*/
if (strncmp(v, "TRUE", 4) == 0) {
DEBUG(2, (" object %s is recycled, not returning linked attribute !\n",
ldb_dn_get_linearized(msg->dn)));
return WERR_OK;
}
} else {
return WERR_OK;
}
}
status = dsdb_get_extended_dn_guid(dsdb_dn->dn, &guid, "GUID");
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ " Unable to extract GUID in linked attribute '%s' in '%s'\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
return ntstatus_to_werror(status);
}
ret = dsdb_find_dn_by_guid(sam_ctx, mem_ctx, &guid, 0, &tdn);
if (ret == LDB_ERR_NO_SUCH_OBJECT) {
DEBUG(2, (" Search of guid %s returned 0 objects, skipping it !\n",
GUID_string(mem_ctx, &guid)));
return WERR_OK;
} else if (ret != LDB_SUCCESS) {
DEBUG(0, (__location__ " Search of guid %s failed with error code %d\n",
GUID_string(mem_ctx, &guid),
ret));
return WERR_OK;
}
}
la->attid = dsdb_attribute_get_attid(sa, is_schema_nc);
la->flags = active?DRSUAPI_DS_LINKED_ATTRIBUTE_FLAG_ACTIVE:0;
status = dsdb_get_extended_dn_uint32(dsdb_dn->dn, &la->meta_data.version, "RMD_VERSION");
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ " No RMD_VERSION in linked attribute '%s' in '%s'\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
return ntstatus_to_werror(status);
}
status = dsdb_get_extended_dn_nttime(dsdb_dn->dn, &la->meta_data.originating_change_time, "RMD_CHANGETIME");
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ " No RMD_CHANGETIME in linked attribute '%s' in '%s'\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
return ntstatus_to_werror(status);
}
status = dsdb_get_extended_dn_guid(dsdb_dn->dn, &la->meta_data.originating_invocation_id, "RMD_INVOCID");
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ " No RMD_INVOCID in linked attribute '%s' in '%s'\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
return ntstatus_to_werror(status);
}
status = dsdb_get_extended_dn_uint64(dsdb_dn->dn, &la->meta_data.originating_usn, "RMD_ORIGINATING_USN");
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ " No RMD_ORIGINATING_USN in linked attribute '%s' in '%s'\n",
sa->lDAPDisplayName, ldb_dn_get_linearized(msg->dn)));
return ntstatus_to_werror(status);
}
status = dsdb_get_extended_dn_nttime(dsdb_dn->dn, &la->originating_add_time, "RMD_ADDTIME");
if (!NT_STATUS_IS_OK(status)) {
/* this is possible for upgraded links */
la->originating_add_time = la->meta_data.originating_change_time;
}
werr = dsdb_dn_la_to_blob(sam_ctx, sa, schema, *la_list, dsdb_dn, &la->value.blob);
W_ERROR_NOT_OK_RETURN(werr);
(*la_count)++;
return WERR_OK;
}
/*
add linked attributes from an object to the list of linked
attributes in a getncchanges request
*/
static WERROR get_nc_changes_add_links(struct ldb_context *sam_ctx,
TALLOC_CTX *mem_ctx,
bool is_schema_nc,
struct dsdb_schema *schema,
uint64_t highest_usn,
uint32_t replica_flags,
const struct ldb_message *msg,
struct drsuapi_DsReplicaLinkedAttribute **la_list,
uint32_t *la_count,
struct drsuapi_DsReplicaCursorCtrEx *uptodateness_vector)
{
unsigned int i;
TALLOC_CTX *tmp_ctx = NULL;
uint64_t uSNChanged = ldb_msg_find_attr_as_uint64(msg, "uSNChanged", 0);
bool is_critical = ldb_msg_find_attr_as_bool(msg, "isCriticalSystemObject", false);
if (replica_flags & DRSUAPI_DRS_CRITICAL_ONLY) {
if (!is_critical) {
return WERR_OK;
}
}
if (uSNChanged <= highest_usn) {
return WERR_OK;
}
tmp_ctx = talloc_new(mem_ctx);
if (tmp_ctx == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
for (i=0; i<msg->num_elements; i++) {
struct ldb_message_element *el = &msg->elements[i];
const struct dsdb_attribute *sa;
unsigned int j;
sa = dsdb_attribute_by_lDAPDisplayName(schema, el->name);
if (!sa || sa->linkID == 0 || (sa->linkID & 1)) {
/* we only want forward links */
continue;
}
if (el->num_values && !dsdb_dn_is_upgraded_link_val(&el->values[0])) {
/* its an old style link, it will have been
* sent in the main replication data */
continue;
}
for (j=0; j<el->num_values; j++) {
struct dsdb_dn *dsdb_dn;
uint64_t local_usn;
uint64_t originating_usn;
NTSTATUS status, status2;
WERROR werr;
struct GUID originating_invocation_id;
dsdb_dn = dsdb_dn_parse(tmp_ctx, sam_ctx, &el->values[j], sa->syntax->ldap_oid);
if (dsdb_dn == NULL) {
DEBUG(1,(__location__ ": Failed to parse DN for %s in %s\n",
el->name, ldb_dn_get_linearized(msg->dn)));
talloc_free(tmp_ctx);
return WERR_DS_DRA_INTERNAL_ERROR;
}
status = dsdb_get_extended_dn_uint64(dsdb_dn->dn, &local_usn, "RMD_LOCAL_USN");
if (!NT_STATUS_IS_OK(status)) {
/* this can happen for attributes
given to us with old style meta
data */
continue;
}
if (local_usn > uSNChanged) {
DEBUG(1,(__location__ ": uSNChanged less than RMD_LOCAL_USN for %s on %s\n",
el->name, ldb_dn_get_linearized(msg->dn)));
talloc_free(tmp_ctx);
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (local_usn <= highest_usn) {
continue;
}
status = dsdb_get_extended_dn_guid(dsdb_dn->dn,
&originating_invocation_id,
"RMD_INVOCID");
status2 = dsdb_get_extended_dn_uint64(dsdb_dn->dn,
&originating_usn,
"RMD_ORIGINATING_USN");
if (NT_STATUS_IS_OK(status) && NT_STATUS_IS_OK(status2)) {
if (udv_filter(uptodateness_vector,
&originating_invocation_id,
originating_usn)) {
continue;
}
}
werr = get_nc_changes_add_la(mem_ctx, sam_ctx, schema,
sa, msg, dsdb_dn, la_list,
la_count, is_schema_nc);
if (!W_ERROR_IS_OK(werr)) {
talloc_free(tmp_ctx);
return werr;
}
}
}
talloc_free(tmp_ctx);
return WERR_OK;
}
/*
fill in the cursors return based on the replUpToDateVector for the ncRoot_dn
*/
static WERROR get_nc_changes_udv(struct ldb_context *sam_ctx,
struct ldb_dn *ncRoot_dn,
struct drsuapi_DsReplicaCursor2CtrEx *udv)
{
int ret;
udv->version = 2;
udv->reserved1 = 0;
udv->reserved2 = 0;
ret = dsdb_load_udv_v2(sam_ctx, ncRoot_dn, udv, &udv->cursors, &udv->count);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed to load UDV for %s - %s\n",
ldb_dn_get_linearized(ncRoot_dn), ldb_errstring(sam_ctx)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
2009-09-10 13:41:48 +04:00
return WERR_OK;
}
/* comparison function for linked attributes - see CompareLinks() in
* MS-DRSR section 4.1.10.5.17 */
static int linked_attribute_compare(const struct la_for_sorting *la1,
const struct la_for_sorting *la2)
{
int c;
c = memcmp(la1->source_guid,
la2->source_guid, sizeof(la2->source_guid));
if (c != 0) {
return c;
}
if (la1->link->attid != la2->link->attid) {
return la1->link->attid < la2->link->attid? -1:1;
}
if ((la1->link->flags & DRSUAPI_DS_LINKED_ATTRIBUTE_FLAG_ACTIVE) !=
(la2->link->flags & DRSUAPI_DS_LINKED_ATTRIBUTE_FLAG_ACTIVE)) {
return (la1->link->flags &
DRSUAPI_DS_LINKED_ATTRIBUTE_FLAG_ACTIVE)? 1:-1;
}
return memcmp(la1->target_guid,
la2->target_guid, sizeof(la2->target_guid));
}
struct drsuapi_changed_objects {
struct ldb_dn *dn;
struct GUID guid;
uint64_t usn;
};
/*
sort the objects we send first by uSNChanged
*/
static int site_res_cmp_usn_order(struct drsuapi_changed_objects *m1,
struct drsuapi_changed_objects *m2,
struct drsuapi_getncchanges_state *getnc_state)
{
int ret;
ret = ldb_dn_compare(getnc_state->ncRoot_dn, m1->dn);
if (ret == 0) {
return -1;
}
ret = ldb_dn_compare(getnc_state->ncRoot_dn, m2->dn);
if (ret == 0) {
return 1;
}
if (m1->usn == m2->usn) {
return ldb_dn_compare(m2->dn, m1->dn);
}
if (m1->usn < m2->usn) {
return -1;
}
return 1;
}
/*
handle a DRSUAPI_EXOP_FSMO_RID_ALLOC call
*/
static WERROR getncchanges_rid_alloc(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct drsuapi_DsGetNCChangesCtr6 *ctr6,
struct ldb_dn **rid_manager_dn)
{
struct ldb_dn *req_dn, *ntds_dn = NULL;
int ret;
struct ldb_context *ldb = b_state->sam_ctx;
struct ldb_result *ext_res;
struct dsdb_fsmo_extended_op *exop;
bool is_us;
/*
steps:
- verify that the DN being asked for is the RID Manager DN
- verify that we are the RID Manager
*/
/* work out who is the RID Manager, also return to caller */
ret = samdb_rid_manager_dn(ldb, mem_ctx, rid_manager_dn);
if (ret != LDB_SUCCESS) {
DEBUG(0, (__location__ ": Failed to find RID Manager object - %s\n", ldb_errstring(ldb)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
req_dn = drs_ObjectIdentifier_to_dn(mem_ctx, ldb, req10->naming_context);
if (!ldb_dn_validate(req_dn) ||
ldb_dn_compare(req_dn, *rid_manager_dn) != 0) {
/* that isn't the RID Manager DN */
DEBUG(0,(__location__ ": RID Alloc request for wrong DN %s\n",
drs_ObjectIdentifier_to_string(mem_ctx, req10->naming_context)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_MISMATCH;
return WERR_OK;
}
/* TODO: make sure ntds_dn is a valid nTDSDSA object */
ret = dsdb_find_dn_by_guid(ldb, mem_ctx, &req10->destination_dsa_guid, 0, &ntds_dn);
if (ret != LDB_SUCCESS) {
DEBUG(0, (__location__ ": Unable to find NTDS object for guid %s - %s\n",
GUID_string(mem_ctx, &req10->destination_dsa_guid), ldb_errstring(ldb)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_UNKNOWN_CALLER;
return WERR_OK;
}
/* find the DN of the RID Manager */
ret = samdb_reference_dn_is_our_ntdsa(ldb, *rid_manager_dn, "fSMORoleOwner", &is_us);
if (ret != LDB_SUCCESS) {
DEBUG(0,("Failed to find fSMORoleOwner in RID Manager object\n"));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_FSMO_NOT_OWNER;
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (!is_us) {
/* we're not the RID Manager - go away */
DEBUG(0,(__location__ ": RID Alloc request when not RID Manager\n"));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_FSMO_NOT_OWNER;
return WERR_OK;
}
exop = talloc(mem_ctx, struct dsdb_fsmo_extended_op);
W_ERROR_HAVE_NO_MEMORY(exop);
exop->fsmo_info = req10->fsmo_info;
exop->destination_dsa_guid = req10->destination_dsa_guid;
ret = ldb_transaction_start(ldb);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction start - %s\n",
ldb_errstring(ldb)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = ldb_extended(ldb, DSDB_EXTENDED_ALLOCATE_RID_POOL, exop, &ext_res);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed extended allocation RID pool operation - %s\n",
ldb_errstring(ldb)));
ldb_transaction_cancel(ldb);
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = ldb_transaction_commit(ldb);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction commit - %s\n",
ldb_errstring(ldb)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
talloc_free(ext_res);
DEBUG(2,("Allocated RID pool for server %s\n",
GUID_string(mem_ctx, &req10->destination_dsa_guid)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_SUCCESS;
return WERR_OK;
}
/*
handle a DRSUAPI_EXOP_REPL_SECRET call
*/
static WERROR getncchanges_repl_secret(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct dom_sid *user_sid,
struct drsuapi_DsGetNCChangesCtr6 *ctr6,
bool has_get_all_changes,
struct ldb_dn **machine_dn)
{
struct drsuapi_DsReplicaObjectIdentifier *ncRoot = req10->naming_context;
struct ldb_dn *obj_dn = NULL;
struct ldb_dn *ntds_dn = NULL, *server_dn = NULL;
struct ldb_dn *rodc_dn, *krbtgt_link_dn;
int ret;
const char *rodc_attrs[] = { "msDS-KrbTgtLink", "msDS-NeverRevealGroup", "msDS-RevealOnDemandGroup", "objectGUID", NULL };
const char *obj_attrs[] = { "tokenGroups", "objectSid", "UserAccountControl", "msDS-KrbTgtLinkBL", NULL };
struct ldb_result *rodc_res = NULL, *obj_res = NULL;
const struct dom_sid **never_reveal_sids, **reveal_sids, **token_sids;
const struct dom_sid *object_sid = NULL;
WERROR werr;
const struct dom_sid *additional_sids[] = { NULL, NULL };
DEBUG(3,(__location__ ": DRSUAPI_EXOP_REPL_SECRET extended op on %s\n",
drs_ObjectIdentifier_to_string(mem_ctx, ncRoot)));
/*
* we need to work out if we will allow this DC to
* replicate the secrets for this object
*
* see 4.1.10.5.14 GetRevealSecretsPolicyForUser for details
* of this function
*/
if (b_state->sam_ctx_system == NULL) {
/* this operation needs system level access */
ctr6->extended_ret = DRSUAPI_EXOP_ERR_ACCESS_DENIED;
return WERR_DS_DRA_ACCESS_DENIED;
}
/*
* Before we accept or deny, fetch the machine DN for the destination
* DSA GUID.
*
* If we are the RODC, we will check that this matches the SID.
*/
ret = dsdb_find_dn_by_guid(b_state->sam_ctx_system, mem_ctx,
&req10->destination_dsa_guid, 0,
&ntds_dn);
if (ret != LDB_SUCCESS) {
goto failed;
}
server_dn = ldb_dn_get_parent(mem_ctx, ntds_dn);
if (server_dn == NULL) {
goto failed;
}
ret = samdb_reference_dn(b_state->sam_ctx_system, mem_ctx, server_dn,
"serverReference", machine_dn);
if (ret != LDB_SUCCESS) {
goto failed;
}
/*
* In MS-DRSR.pdf 5.99 IsGetNCChangesPermissionGranted
*
* The pseudo code indicate
* revealsecrets = true
* if IsRevealSecretRequest(msgIn) then
* if AccessCheckCAR(ncRoot, Ds-Replication-Get-Changes-All) = false
* then
* if (msgIn.ulExtendedOp = EXOP_REPL_SECRETS) then
* <... check if this account is ok to be replicated on this DC ...>
* <... and if not reveal secrets = no ...>
* else
* reveal secrets = false
* endif
* endif
* endif
*
* Which basically means that if you have GET_ALL_CHANGES rights (~== RWDC)
* then you can do EXOP_REPL_SECRETS
*/
obj_dn = drs_ObjectIdentifier_to_dn(mem_ctx, b_state->sam_ctx_system, ncRoot);
if (!ldb_dn_validate(obj_dn)) goto failed;
if (has_get_all_changes) {
goto allowed;
}
rodc_dn = ldb_dn_new_fmt(mem_ctx, b_state->sam_ctx_system, "<SID=%s>",
dom_sid_string(mem_ctx, user_sid));
if (!ldb_dn_validate(rodc_dn)) goto failed;
/* do the two searches we need */
ret = dsdb_search_dn(b_state->sam_ctx_system, mem_ctx, &rodc_res, rodc_dn, rodc_attrs,
DSDB_SEARCH_SHOW_EXTENDED_DN);
if (ret != LDB_SUCCESS || rodc_res->count != 1) goto failed;
ret = dsdb_search_dn(b_state->sam_ctx_system, mem_ctx, &obj_res, obj_dn, obj_attrs, 0);
if (ret != LDB_SUCCESS || obj_res->count != 1) goto failed;
/* if the object SID is equal to the user_sid, allow */
object_sid = samdb_result_dom_sid(mem_ctx, obj_res->msgs[0], "objectSid");
if (dom_sid_equal(user_sid, object_sid)) {
goto allowed;
}
additional_sids[0] = object_sid;
/*
* Must be an RODC account at this point, verify machine DN matches the
* SID account
*/
if (ldb_dn_compare(rodc_res->msgs[0]->dn, *machine_dn) != 0) {
goto denied;
}
/* an RODC is allowed to get its own krbtgt account secrets */
krbtgt_link_dn = samdb_result_dn(b_state->sam_ctx_system, mem_ctx,
rodc_res->msgs[0], "msDS-KrbTgtLink", NULL);
if (krbtgt_link_dn != NULL &&
ldb_dn_compare(obj_dn, krbtgt_link_dn) == 0) {
goto allowed;
}
/* but it isn't allowed to get anyone elses krbtgt secrets */
if (samdb_result_dn(b_state->sam_ctx_system, mem_ctx,
obj_res->msgs[0], "msDS-KrbTgtLinkBL", NULL)) {
goto denied;
}
if (ldb_msg_find_attr_as_uint(obj_res->msgs[0],
"userAccountControl", 0) &
UF_INTERDOMAIN_TRUST_ACCOUNT) {
goto denied;
}
werr = samdb_result_sid_array_dn(b_state->sam_ctx_system, rodc_res->msgs[0],
mem_ctx, "msDS-NeverRevealGroup", &never_reveal_sids);
if (!W_ERROR_IS_OK(werr)) {
goto denied;
}
werr = samdb_result_sid_array_dn(b_state->sam_ctx_system, rodc_res->msgs[0],
mem_ctx, "msDS-RevealOnDemandGroup", &reveal_sids);
if (!W_ERROR_IS_OK(werr)) {
goto denied;
}
/*
* The SID list needs to include itself as well as the tokenGroups.
*
* TODO determine if sIDHistory is required for this check
*/
werr = samdb_result_sid_array_ndr(b_state->sam_ctx_system, obj_res->msgs[0],
mem_ctx, "tokenGroups", &token_sids,
additional_sids, 1);
if (!W_ERROR_IS_OK(werr) || token_sids==NULL) {
goto denied;
}
if (never_reveal_sids &&
sid_list_match(token_sids, never_reveal_sids)) {
goto denied;
}
if (reveal_sids &&
sid_list_match(token_sids, reveal_sids)) {
goto allowed;
}
/* default deny */
denied:
DEBUG(2,(__location__ ": Denied single object with secret replication for %s by RODC %s\n",
ldb_dn_get_linearized(obj_dn), ldb_dn_get_linearized(rodc_res->msgs[0]->dn)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_NONE;
return WERR_DS_DRA_SECRETS_DENIED;
allowed:
DEBUG(2,(__location__ ": Allowed single object with secret replication for %s by %s %s\n",
ldb_dn_get_linearized(obj_dn), has_get_all_changes?"RWDC":"RODC",
ldb_dn_get_linearized(*machine_dn)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_SUCCESS;
req10->highwatermark.highest_usn = 0;
return WERR_OK;
failed:
DEBUG(2,(__location__ ": Failed single secret replication for %s by RODC %s\n",
ldb_dn_get_linearized(obj_dn), dom_sid_string(mem_ctx, user_sid)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_NONE;
return WERR_DS_DRA_BAD_DN;
}
/*
handle a DRSUAPI_EXOP_REPL_OBJ call
*/
static WERROR getncchanges_repl_obj(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct dom_sid *user_sid,
struct drsuapi_DsGetNCChangesCtr6 *ctr6)
{
struct drsuapi_DsReplicaObjectIdentifier *ncRoot = req10->naming_context;
DEBUG(3,(__location__ ": DRSUAPI_EXOP_REPL_OBJ extended op on %s\n",
drs_ObjectIdentifier_to_string(mem_ctx, ncRoot)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_SUCCESS;
return WERR_OK;
}
/*
handle DRSUAPI_EXOP_FSMO_REQ_ROLE,
DRSUAPI_EXOP_FSMO_RID_REQ_ROLE,
and DRSUAPI_EXOP_FSMO_REQ_PDC calls
*/
static WERROR getncchanges_change_master(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct drsuapi_DsGetNCChangesCtr6 *ctr6)
{
struct ldb_dn *req_dn, *ntds_dn;
int ret;
unsigned int i;
struct ldb_context *ldb = b_state->sam_ctx;
struct ldb_message *msg;
bool is_us;
/*
steps:
- verify that the client dn exists
- verify that we are the current master
*/
req_dn = drs_ObjectIdentifier_to_dn(mem_ctx, ldb, req10->naming_context);
if (!ldb_dn_validate(req_dn)) {
/* that is not a valid dn */
DEBUG(0,(__location__ ": FSMO role transfer request for invalid DN %s\n",
drs_ObjectIdentifier_to_string(mem_ctx, req10->naming_context)));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_MISMATCH;
return WERR_OK;
}
/* find the DN of the current role owner */
ret = samdb_reference_dn_is_our_ntdsa(ldb, req_dn, "fSMORoleOwner", &is_us);
if (ret != LDB_SUCCESS) {
DEBUG(0,("Failed to find fSMORoleOwner in RID Manager object\n"));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_FSMO_NOT_OWNER;
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (!is_us) {
/* we're not the RID Manager or role owner - go away */
DEBUG(0,(__location__ ": FSMO role or RID manager transfer owner request when not role owner\n"));
ctr6->extended_ret = DRSUAPI_EXOP_ERR_FSMO_NOT_OWNER;
return WERR_OK;
}
/* change the current master */
msg = ldb_msg_new(ldb);
W_ERROR_HAVE_NO_MEMORY(msg);
msg->dn = drs_ObjectIdentifier_to_dn(msg, ldb, req10->naming_context);
W_ERROR_HAVE_NO_MEMORY(msg->dn);
/* TODO: make sure ntds_dn is a valid nTDSDSA object */
ret = dsdb_find_dn_by_guid(ldb, msg, &req10->destination_dsa_guid, 0, &ntds_dn);
if (ret != LDB_SUCCESS) {
DEBUG(0, (__location__ ": Unable to find NTDS object for guid %s - %s\n",
GUID_string(mem_ctx, &req10->destination_dsa_guid), ldb_errstring(ldb)));
talloc_free(msg);
ctr6->extended_ret = DRSUAPI_EXOP_ERR_UNKNOWN_CALLER;
return WERR_OK;
}
ret = ldb_msg_add_string(msg, "fSMORoleOwner", ldb_dn_get_linearized(ntds_dn));
if (ret != 0) {
talloc_free(msg);
return WERR_DS_DRA_INTERNAL_ERROR;
}
for (i=0;i<msg->num_elements;i++) {
msg->elements[i].flags = LDB_FLAG_MOD_REPLACE;
}
ret = ldb_transaction_start(ldb);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction start - %s\n",
ldb_errstring(ldb)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = ldb_modify(ldb, msg);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed to change current owner - %s\n",
ldb_errstring(ldb)));
ldb_transaction_cancel(ldb);
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = ldb_transaction_commit(ldb);
if (ret != LDB_SUCCESS) {
DEBUG(0,(__location__ ": Failed transaction commit - %s\n",
ldb_errstring(ldb)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
ctr6->extended_ret = DRSUAPI_EXOP_ERR_SUCCESS;
return WERR_OK;
}
/*
see if this getncchanges request includes a request to reveal secret information
*/
static WERROR dcesrv_drsuapi_is_reveal_secrets_request(struct drsuapi_bind_state *b_state,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct dsdb_schema_prefixmap *pfm_remote,
bool *is_secret_request)
{
enum drsuapi_DsExtendedOperation exop;
uint32_t i;
struct dsdb_schema *schema;
struct dsdb_syntax_ctx syntax_ctx;
*is_secret_request = true;
exop = req10->extended_op;
switch (exop) {
case DRSUAPI_EXOP_FSMO_REQ_ROLE:
case DRSUAPI_EXOP_FSMO_RID_ALLOC:
case DRSUAPI_EXOP_FSMO_RID_REQ_ROLE:
case DRSUAPI_EXOP_FSMO_REQ_PDC:
case DRSUAPI_EXOP_FSMO_ABANDON_ROLE:
/* FSMO exops can reveal secrets */
*is_secret_request = true;
return WERR_OK;
case DRSUAPI_EXOP_REPL_SECRET:
case DRSUAPI_EXOP_REPL_OBJ:
case DRSUAPI_EXOP_NONE:
break;
}
if (req10->replica_flags & DRSUAPI_DRS_SPECIAL_SECRET_PROCESSING) {
*is_secret_request = false;
return WERR_OK;
}
if (exop == DRSUAPI_EXOP_REPL_SECRET ||
req10->partial_attribute_set == NULL) {
/* they want secrets */
*is_secret_request = true;
return WERR_OK;
}
schema = dsdb_get_schema(b_state->sam_ctx, NULL);
dsdb_syntax_ctx_init(&syntax_ctx, b_state->sam_ctx, schema);
syntax_ctx.pfm_remote = pfm_remote;
/* check the attributes they asked for */
for (i=0; i<req10->partial_attribute_set->num_attids; i++) {
const struct dsdb_attribute *sa;
WERROR werr = getncchanges_attid_remote_to_local(schema,
&syntax_ctx,
req10->partial_attribute_set->attids[i],
NULL,
&sa);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__": attid 0x%08X not found: %s\n",
req10->partial_attribute_set->attids[i], win_errstr(werr)));
return werr;
}
if (!dsdb_attr_in_rodc_fas(sa)) {
*is_secret_request = true;
return WERR_OK;
}
}
if (req10->partial_attribute_set_ex) {
/* check the extended attributes they asked for */
for (i=0; i<req10->partial_attribute_set_ex->num_attids; i++) {
const struct dsdb_attribute *sa;
WERROR werr = getncchanges_attid_remote_to_local(schema,
&syntax_ctx,
req10->partial_attribute_set_ex->attids[i],
NULL,
&sa);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__": attid 0x%08X not found: %s\n",
req10->partial_attribute_set_ex->attids[i], win_errstr(werr)));
return werr;
}
if (!dsdb_attr_in_rodc_fas(sa)) {
*is_secret_request = true;
return WERR_OK;
}
}
}
*is_secret_request = false;
return WERR_OK;
}
/*
see if this getncchanges request is only for attributes in the GC
partial attribute set
*/
static WERROR dcesrv_drsuapi_is_gc_pas_request(struct drsuapi_bind_state *b_state,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct dsdb_schema_prefixmap *pfm_remote,
bool *is_gc_pas_request)
{
enum drsuapi_DsExtendedOperation exop;
uint32_t i;
struct dsdb_schema *schema;
struct dsdb_syntax_ctx syntax_ctx;
exop = req10->extended_op;
switch (exop) {
case DRSUAPI_EXOP_FSMO_REQ_ROLE:
case DRSUAPI_EXOP_FSMO_RID_ALLOC:
case DRSUAPI_EXOP_FSMO_RID_REQ_ROLE:
case DRSUAPI_EXOP_FSMO_REQ_PDC:
case DRSUAPI_EXOP_FSMO_ABANDON_ROLE:
case DRSUAPI_EXOP_REPL_SECRET:
*is_gc_pas_request = false;
return WERR_OK;
case DRSUAPI_EXOP_REPL_OBJ:
case DRSUAPI_EXOP_NONE:
break;
}
if (req10->partial_attribute_set == NULL) {
/* they want it all */
*is_gc_pas_request = false;
return WERR_OK;
}
schema = dsdb_get_schema(b_state->sam_ctx, NULL);
dsdb_syntax_ctx_init(&syntax_ctx, b_state->sam_ctx, schema);
syntax_ctx.pfm_remote = pfm_remote;
/* check the attributes they asked for */
for (i=0; i<req10->partial_attribute_set->num_attids; i++) {
const struct dsdb_attribute *sa;
WERROR werr = getncchanges_attid_remote_to_local(schema,
&syntax_ctx,
req10->partial_attribute_set->attids[i],
NULL,
&sa);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__": attid 0x%08X not found: %s\n",
req10->partial_attribute_set->attids[i], win_errstr(werr)));
return werr;
}
if (!sa->isMemberOfPartialAttributeSet) {
*is_gc_pas_request = false;
return WERR_OK;
}
}
if (req10->partial_attribute_set_ex) {
/* check the extended attributes they asked for */
for (i=0; i<req10->partial_attribute_set_ex->num_attids; i++) {
const struct dsdb_attribute *sa;
WERROR werr = getncchanges_attid_remote_to_local(schema,
&syntax_ctx,
req10->partial_attribute_set_ex->attids[i],
NULL,
&sa);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__": attid 0x%08X not found: %s\n",
req10->partial_attribute_set_ex->attids[i], win_errstr(werr)));
return werr;
}
if (!sa->isMemberOfPartialAttributeSet) {
*is_gc_pas_request = false;
return WERR_OK;
}
}
}
*is_gc_pas_request = true;
return WERR_OK;
}
/*
map from req8 to req10
*/
static struct drsuapi_DsGetNCChangesRequest10 *
getncchanges_map_req8(TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest8 *req8)
{
struct drsuapi_DsGetNCChangesRequest10 *req10 = talloc_zero(mem_ctx,
struct drsuapi_DsGetNCChangesRequest10);
if (req10 == NULL) {
return NULL;
}
req10->destination_dsa_guid = req8->destination_dsa_guid;
req10->source_dsa_invocation_id = req8->source_dsa_invocation_id;
req10->naming_context = req8->naming_context;
req10->highwatermark = req8->highwatermark;
req10->uptodateness_vector = req8->uptodateness_vector;
req10->replica_flags = req8->replica_flags;
req10->max_object_count = req8->max_object_count;
req10->max_ndr_size = req8->max_ndr_size;
req10->extended_op = req8->extended_op;
req10->fsmo_info = req8->fsmo_info;
req10->partial_attribute_set = req8->partial_attribute_set;
req10->partial_attribute_set_ex = req8->partial_attribute_set_ex;
req10->mapping_ctr = req8->mapping_ctr;
return req10;
}
static const char *collect_objects_attrs[] = { "uSNChanged",
"objectGUID" ,
NULL };
/**
* Collects object for normal replication cycle.
*/
static WERROR getncchanges_collect_objects(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct ldb_dn *search_dn,
const char *extra_filter,
struct ldb_result **search_res)
{
int ret;
char* search_filter;
enum ldb_scope scope = LDB_SCOPE_SUBTREE;
struct drsuapi_getncchanges_state *getnc_state = b_state->getncchanges_state;
bool critical_only = false;
if (req10->replica_flags & DRSUAPI_DRS_CRITICAL_ONLY) {
critical_only = true;
}
if (req10->extended_op == DRSUAPI_EXOP_REPL_OBJ ||
req10->extended_op == DRSUAPI_EXOP_REPL_SECRET) {
scope = LDB_SCOPE_BASE;
critical_only = false;
}
/* Construct response. */
search_filter = talloc_asprintf(mem_ctx,
"(uSNChanged>=%llu)",
(unsigned long long)(getnc_state->min_usn+1));
if (extra_filter) {
search_filter = talloc_asprintf(mem_ctx, "(&%s(%s))", search_filter, extra_filter);
}
if (critical_only) {
search_filter = talloc_asprintf(mem_ctx,
"(&%s(isCriticalSystemObject=TRUE))",
search_filter);
}
if (req10->replica_flags & DRSUAPI_DRS_ASYNC_REP) {
scope = LDB_SCOPE_BASE;
}
if (!search_dn) {
search_dn = getnc_state->ncRoot_dn;
}
DEBUG(2,(__location__ ": getncchanges on %s using filter %s\n",
ldb_dn_get_linearized(getnc_state->ncRoot_dn), search_filter));
ret = drsuapi_search_with_extended_dn(b_state->sam_ctx, getnc_state, search_res,
search_dn, scope,
collect_objects_attrs,
search_filter);
if (ret != LDB_SUCCESS) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
return WERR_OK;
}
/**
* Collects object for normal replication cycle.
*/
static WERROR getncchanges_collect_objects_exop(struct drsuapi_bind_state *b_state,
TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChangesRequest10 *req10,
struct drsuapi_DsGetNCChangesCtr6 *ctr6,
struct ldb_dn *search_dn,
const char *extra_filter,
struct ldb_result **search_res)
{
/* we have nothing to do in case of ex-op failure */
if (ctr6->extended_ret != DRSUAPI_EXOP_ERR_SUCCESS) {
return WERR_OK;
}
switch (req10->extended_op) {
case DRSUAPI_EXOP_FSMO_RID_ALLOC:
{
int ret;
struct ldb_dn *ntds_dn = NULL;
struct ldb_dn *server_dn = NULL;
struct ldb_dn *machine_dn = NULL;
struct ldb_dn *rid_set_dn = NULL;
struct ldb_result *search_res2 = NULL;
struct ldb_result *search_res3 = NULL;
TALLOC_CTX *frame = talloc_stackframe();
/* get RID manager, RID set and server DN (in that order) */
/* This first search will get the RID Manager */
ret = drsuapi_search_with_extended_dn(b_state->sam_ctx, frame,
search_res,
search_dn, LDB_SCOPE_BASE,
collect_objects_attrs,
NULL);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get RID Manager object %s - %s",
ldb_dn_get_linearized(search_dn),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
if ((*search_res)->count != 1) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get RID Manager object %s - %u objects returned",
ldb_dn_get_linearized(search_dn),
(*search_res)->count));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
/* Now extend it to the RID set */
/* Find the computer account DN for the destination
* dsa GUID specified */
ret = dsdb_find_dn_by_guid(b_state->sam_ctx, frame,
&req10->destination_dsa_guid, 0,
&ntds_dn);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Unable to find NTDS object for guid %s - %s\n",
GUID_string(frame,
&req10->destination_dsa_guid),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
server_dn = ldb_dn_get_parent(frame, ntds_dn);
if (!server_dn) {
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = samdb_reference_dn(b_state->sam_ctx, frame, server_dn,
"serverReference", &machine_dn);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to find serverReference in %s - %s",
ldb_dn_get_linearized(server_dn),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
ret = samdb_reference_dn(b_state->sam_ctx, frame, machine_dn,
"rIDSetReferences", &rid_set_dn);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to find rIDSetReferences in %s - %s",
ldb_dn_get_linearized(server_dn),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
/* This first search will get the RID Manager, now get the RID set */
ret = drsuapi_search_with_extended_dn(b_state->sam_ctx, frame,
&search_res2,
rid_set_dn, LDB_SCOPE_BASE,
collect_objects_attrs,
NULL);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get RID Set object %s - %s",
ldb_dn_get_linearized(rid_set_dn),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (search_res2->count != 1) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get RID Set object %s - %u objects returned",
ldb_dn_get_linearized(rid_set_dn),
search_res2->count));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
/* Finally get the server DN */
ret = drsuapi_search_with_extended_dn(b_state->sam_ctx, frame,
&search_res3,
machine_dn, LDB_SCOPE_BASE,
collect_objects_attrs,
NULL);
if (ret != LDB_SUCCESS) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get server object %s - %s",
ldb_dn_get_linearized(server_dn),
ldb_errstring(b_state->sam_ctx)));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
if (search_res3->count != 1) {
DEBUG(1, ("DRSUAPI_EXOP_FSMO_RID_ALLOC: Failed to get server object %s - %u objects returned",
ldb_dn_get_linearized(server_dn),
search_res3->count));
TALLOC_FREE(frame);
return WERR_DS_DRA_INTERNAL_ERROR;
}
/* Now extend the original search_res with these answers */
(*search_res)->count = 3;
(*search_res)->msgs = talloc_realloc(frame, (*search_res)->msgs,
struct ldb_message *,
(*search_res)->count);
if ((*search_res)->msgs == NULL) {
TALLOC_FREE(frame);
return WERR_NOT_ENOUGH_MEMORY;
}
talloc_steal(mem_ctx, *search_res);
(*search_res)->msgs[1] =
talloc_steal((*search_res)->msgs, search_res2->msgs[0]);
(*search_res)->msgs[2] =
talloc_steal((*search_res)->msgs, search_res3->msgs[0]);
TALLOC_FREE(frame);
return WERR_OK;
}
default:
/* TODO: implement extended op specific collection
* of objects. Right now we just normal procedure
* for collecting objects */
return getncchanges_collect_objects(b_state, mem_ctx, req10, search_dn, extra_filter, search_res);
}
}
static void dcesrv_drsuapi_update_highwatermark(const struct ldb_message *msg,
uint64_t max_usn,
struct drsuapi_DsReplicaHighWaterMark *hwm)
{
uint64_t uSN = ldb_msg_find_attr_as_uint64(msg, "uSNChanged", 0);
if (uSN > max_usn) {
/*
* Only report the max_usn we had at the start
* of the replication cycle.
*
* If this object has changed lately we better
* let the destination dsa refetch the change.
* This is better than the risk of loosing some
* objects or linked attributes.
*/
return;
}
if (uSN <= hwm->tmp_highest_usn) {
return;
}
hwm->tmp_highest_usn = uSN;
hwm->reserved_usn = 0;
}
/**
* Adds an object's GUID to the cache of objects already sent.
* This avoids us sending the same object multiple times when
* the GetNCChanges request uses a flag like GET_ANC.
*/
static WERROR dcesrv_drsuapi_obj_cache_add(struct db_context *obj_cache,
const struct GUID *guid)
{
enum ndr_err_code ndr_err;
uint8_t guid_buf[DRS_GUID_SIZE] = { 0, };
DATA_BLOB b = {
.data = guid_buf,
.length = sizeof(guid_buf),
};
TDB_DATA key = {
.dptr = b.data,
.dsize = b.length,
};
TDB_DATA val = {
.dptr = NULL,
.dsize = 0,
};
NTSTATUS status;
ndr_err = ndr_push_struct_into_fixed_blob(&b, guid,
(ndr_push_flags_fn_t)ndr_push_GUID);
if (!NDR_ERR_CODE_IS_SUCCESS(ndr_err)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
status = dbwrap_store(obj_cache, key, val, TDB_REPLACE);
if (!NT_STATUS_IS_OK(status)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
return WERR_OK;
}
/**
* Checks if the object with the GUID specified already exists in the
* object cache, i.e. it's already been sent in a GetNCChanges response.
*/
static WERROR dcesrv_drsuapi_obj_cache_exists(struct db_context *obj_cache,
const struct GUID *guid)
{
enum ndr_err_code ndr_err;
uint8_t guid_buf[DRS_GUID_SIZE] = { 0, };
DATA_BLOB b = {
.data = guid_buf,
.length = sizeof(guid_buf),
};
TDB_DATA key = {
.dptr = b.data,
.dsize = b.length,
};
bool exists;
ndr_err = ndr_push_struct_into_fixed_blob(&b, guid,
(ndr_push_flags_fn_t)ndr_push_GUID);
if (!NDR_ERR_CODE_IS_SUCCESS(ndr_err)) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
exists = dbwrap_exists(obj_cache, key);
if (!exists) {
return WERR_OBJECT_NOT_FOUND;
}
return WERR_OBJECT_NAME_EXISTS;
}
/**
* Copies the la_list specified into a sorted array, ready to be sent in a
* GetNCChanges response.
*/
static WERROR getncchanges_get_sorted_array(const struct drsuapi_DsReplicaLinkedAttribute *la_list,
const uint32_t link_count,
struct ldb_context *sam_ctx,
TALLOC_CTX *mem_ctx,
const struct dsdb_schema *schema,
struct la_for_sorting **ret_array)
{
int j;
struct la_for_sorting *guid_array;
WERROR werr = WERR_OK;
*ret_array = NULL;
guid_array = talloc_array(mem_ctx, struct la_for_sorting, link_count);
if (guid_array == NULL) {
DEBUG(0, ("Out of memory allocating %u linked attributes for sorting", link_count));
return WERR_NOT_ENOUGH_MEMORY;
}
for (j = 0; j < link_count; j++) {
/* we need to get the target GUIDs to compare */
struct dsdb_dn *dn;
const struct drsuapi_DsReplicaLinkedAttribute *la = &la_list[j];
const struct dsdb_attribute *schema_attrib;
const struct ldb_val *target_guid;
DATA_BLOB source_guid;
TALLOC_CTX *frame = talloc_stackframe();
NTSTATUS status;
schema_attrib = dsdb_attribute_by_attributeID_id(schema, la->attid);
werr = dsdb_dn_la_from_blob(sam_ctx, schema_attrib, schema, frame, la->value.blob, &dn);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__ ": Bad la blob in sort\n"));
TALLOC_FREE(frame);
return werr;
}
/* Extract the target GUID in NDR form */
target_guid = ldb_dn_get_extended_component(dn->dn, "GUID");
if (target_guid == NULL
|| target_guid->length != sizeof(guid_array[0].target_guid)) {
status = NT_STATUS_OBJECT_NAME_NOT_FOUND;
} else {
/* Repack the source GUID as NDR for sorting */
status = GUID_to_ndr_blob(&la->identifier->guid,
frame,
&source_guid);
}
if (!NT_STATUS_IS_OK(status)
|| source_guid.length != sizeof(guid_array[0].source_guid)) {
DEBUG(0,(__location__ ": Bad la guid in sort\n"));
TALLOC_FREE(frame);
return ntstatus_to_werror(status);
}
guid_array[j].link = &la_list[j];
memcpy(guid_array[j].target_guid, target_guid->data,
sizeof(guid_array[j].target_guid));
memcpy(guid_array[j].source_guid, source_guid.data,
sizeof(guid_array[j].source_guid));
TALLOC_FREE(frame);
}
TYPESAFE_QSORT(guid_array, link_count, linked_attribute_compare);
*ret_array = guid_array;
return werr;
}
/**
* Adds any ancestor/parent objects of the child_obj specified.
* This is needed when the GET_ANC flag is specified in the request.
* @param new_objs if parents are added, this gets updated to point to a chain
* of parent objects (with the parents first and the child last)
*/
static WERROR getncchanges_add_ancestors(struct drsuapi_DsReplicaObjectListItemEx *child_obj,
struct ldb_dn *child_dn,
TALLOC_CTX *mem_ctx,
struct ldb_context *sam_ctx,
struct drsuapi_getncchanges_state *getnc_state,
struct dsdb_schema *schema,
DATA_BLOB *session_key,
struct drsuapi_DsGetNCChangesRequest10 *req10,
uint32_t *local_pas,
struct ldb_dn *machine_dn,
struct drsuapi_DsReplicaObjectListItemEx **new_objs)
{
int ret;
const struct GUID *next_anc_guid = NULL;
WERROR werr = WERR_OK;
static const char * const msg_attrs[] = {
"*",
"nTSecurityDescriptor",
"parentGUID",
"replPropertyMetaData",
DSDB_SECRET_ATTRIBUTES,
NULL };
next_anc_guid = child_obj->parent_object_guid;
while (next_anc_guid != NULL) {
struct drsuapi_DsReplicaObjectListItemEx *anc_obj = NULL;
struct ldb_message *anc_msg = NULL;
struct ldb_result *anc_res = NULL;
struct ldb_dn *anc_dn = NULL;
/*
* Don't send an object twice. (If we've sent the object, then
* we've also sent all its parents as well)
*/
werr = dcesrv_drsuapi_obj_cache_exists(getnc_state->obj_cache,
next_anc_guid);
if (W_ERROR_EQUAL(werr, WERR_OBJECT_NAME_EXISTS)) {
return WERR_OK;
}
if (W_ERROR_IS_OK(werr)) {
return WERR_INTERNAL_ERROR;
}
if (!W_ERROR_EQUAL(werr, WERR_OBJECT_NOT_FOUND)) {
return werr;
}
anc_obj = talloc_zero(mem_ctx,
struct drsuapi_DsReplicaObjectListItemEx);
if (anc_obj == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
anc_dn = ldb_dn_new_fmt(anc_obj, sam_ctx, "<GUID=%s>",
GUID_string(anc_obj, next_anc_guid));
if (anc_dn == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
ret = drsuapi_search_with_extended_dn(sam_ctx, anc_obj,
&anc_res, anc_dn,
LDB_SCOPE_BASE,
msg_attrs, NULL);
if (ret != LDB_SUCCESS) {
const char *anc_str = NULL;
const char *obj_str = NULL;
anc_str = ldb_dn_get_extended_linearized(anc_obj,
anc_dn,
1);
obj_str = ldb_dn_get_extended_linearized(anc_obj,
child_dn,
1);
DBG_ERR("getncchanges: failed to fetch ANC "
"DN %s for DN %s - %s\n",
anc_str, obj_str, ldb_errstring(sam_ctx));
return WERR_DS_DRA_INCONSISTENT_DIT;
}
anc_msg = anc_res->msgs[0];
werr = get_nc_changes_build_object(anc_obj, anc_msg,
sam_ctx,
getnc_state,
schema, session_key,
req10,
false, /* force_object_return */
local_pas,
machine_dn,
next_anc_guid);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
/*
* Regardless of whether we actually use it or not,
* we add it to the cache so we don't look at it again
*/
werr = dcesrv_drsuapi_obj_cache_add(getnc_state->obj_cache,
next_anc_guid);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
/*
* Any ancestors which are below the highwatermark
* or uptodateness_vector shouldn't be added,
* but we still look further up the
* tree for ones which have been changed recently.
*/
if (anc_obj->meta_data_ctr != NULL) {
/*
* prepend the parent to the list so that the client-side
* adds the parent object before it adds the children
*/
anc_obj->next_object = *new_objs;
*new_objs = anc_obj;
}
anc_msg = NULL;
TALLOC_FREE(anc_res);
TALLOC_FREE(anc_dn);
/*
* We may need to resolve more parents...
*/
next_anc_guid = anc_obj->parent_object_guid;
}
return werr;
}
/**
* Adds a list of new objects into the current chunk of replication data to send
*/
static void getncchanges_chunk_add_objects(struct getncchanges_repl_chunk *repl_chunk,
struct drsuapi_DsReplicaObjectListItemEx *obj_list)
{
struct drsuapi_DsReplicaObjectListItemEx *obj;
/*
* We track the last object added to the replication chunk, so just add
* the new object-list onto the end
*/
if (repl_chunk->object_list == NULL) {
repl_chunk->object_list = obj_list;
} else {
repl_chunk->last_object->next_object = obj_list;
}
for (obj = obj_list; obj != NULL; obj = obj->next_object) {
repl_chunk->object_count += 1;
/*
* Remember the last object in the response - we'll use this to
* link the next object(s) processed onto the existing list
*/
if (obj->next_object == NULL) {
repl_chunk->last_object = obj;
}
}
}
/**
* Gets the object to send, packed into an RPC struct ready to send. This also
* adds the object to the object cache, and adds any ancestors (if needed).
* @param msg - DB search result for the object to add
* @param guid - GUID of the object to add
* @param ret_obj_list - returns the object ready to be sent (in a list, along
* with any ancestors that might be needed). NULL if nothing to send.
*/
static WERROR getncchanges_get_obj_to_send(const struct ldb_message *msg,
TALLOC_CTX *mem_ctx,
struct ldb_context *sam_ctx,
struct drsuapi_getncchanges_state *getnc_state,
struct dsdb_schema *schema,
DATA_BLOB *session_key,
struct drsuapi_DsGetNCChangesRequest10 *req10,
bool force_object_return,
uint32_t *local_pas,
struct ldb_dn *machine_dn,
const struct GUID *guid,
struct drsuapi_DsReplicaObjectListItemEx **ret_obj_list)
{
struct drsuapi_DsReplicaObjectListItemEx *obj;
WERROR werr;
*ret_obj_list = NULL;
obj = talloc_zero(mem_ctx, struct drsuapi_DsReplicaObjectListItemEx);
W_ERROR_HAVE_NO_MEMORY(obj);
werr = get_nc_changes_build_object(obj, msg, sam_ctx, getnc_state,
schema, session_key, req10,
force_object_return,
local_pas, machine_dn, guid);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
/*
* The object may get filtered out by the UTDV's USN and not actually
* sent, in which case there's nothing more to do here
*/
if (obj->meta_data_ctr == NULL) {
TALLOC_FREE(obj);
return WERR_OK;
}
if (getnc_state->obj_cache != NULL) {
werr = dcesrv_drsuapi_obj_cache_add(getnc_state->obj_cache,
guid);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
*ret_obj_list = obj;
/*
* If required, also add any ancestors that the client may need to know
* about before it can resolve this object. These get prepended to the
* ret_obj_list so the client adds them first.
*/
if (getnc_state->is_get_anc) {
werr = getncchanges_add_ancestors(obj, msg->dn, mem_ctx,
sam_ctx, getnc_state,
schema, session_key,
req10, local_pas,
machine_dn, ret_obj_list);
}
return werr;
}
/**
* Returns the number of links that are waiting to be sent
*/
static uint32_t getncchanges_chunk_links_pending(struct getncchanges_repl_chunk *repl_chunk,
struct drsuapi_getncchanges_state *getnc_state)
{
uint32_t links_to_send = 0;
if (getnc_state->is_get_tgt) {
/*
* when the GET_TGT flag is set, only include the linked
* attributes whose target object has already been checked
* (i.e. they're ready to send).
*/
if (repl_chunk->tgt_la_count > getnc_state->la_idx) {
links_to_send = (repl_chunk->tgt_la_count -
getnc_state->la_idx);
}
} else {
links_to_send = getnc_state->la_count - getnc_state->la_idx;
}
return links_to_send;
}
/**
* Returns the max number of links that will fit in the current replication chunk
*/
static uint32_t getncchanges_chunk_max_links(struct getncchanges_repl_chunk *repl_chunk)
{
uint32_t max_links = 0;
if (repl_chunk->max_links != DEFAULT_MAX_LINKS ||
repl_chunk->max_objects != DEFAULT_MAX_OBJECTS) {
/*
* We're using non-default settings, so don't try to adjust
* them, just trust the user has configured decent values
*/
max_links = repl_chunk->max_links;
} else if (repl_chunk->max_links > repl_chunk->object_count) {
/*
* This is just an approximate guess to avoid overfilling the
* replication chunk. It's the logic we've used historically.
* E.g. if we've already sent 1000 objects, then send 1000 fewer
* links. For comparison, the max that Windows seems to send is
* ~2700 links and ~250 objects (although this may vary based
* on timeouts)
*/
max_links = repl_chunk->max_links - repl_chunk->object_count;
}
return max_links;
}
/**
* Returns true if the current GetNCChanges() call has taken longer than its
* allotted time. This prevents the client from timing out.
*/
static bool getncchanges_chunk_timed_out(struct getncchanges_repl_chunk *repl_chunk)
{
return (time(NULL) - repl_chunk->start > repl_chunk->max_wait);
}
/**
* Returns true if the current chunk of replication data has reached the
* max_objects and/or max_links thresholds.
*/
static bool getncchanges_chunk_is_full(struct getncchanges_repl_chunk *repl_chunk,
struct drsuapi_getncchanges_state *getnc_state)
{
bool chunk_full = false;
uint32_t links_to_send;
uint32_t chunk_limit;
/* check if the current chunk is already full with objects */
if (repl_chunk->object_count >= repl_chunk->max_objects) {
chunk_full = true;
} else if (repl_chunk->object_count > 0 &&
getncchanges_chunk_timed_out(repl_chunk)) {
/*
* We've exceeded our allotted time building this chunk,
* and we have at least one object to send back to the client
*/
chunk_full = true;
} else if (repl_chunk->immediate_link_sync) {
/* check if the chunk is already full with links */
links_to_send = getncchanges_chunk_links_pending(repl_chunk,
getnc_state);
chunk_limit = getncchanges_chunk_max_links(repl_chunk);
/*
* The chunk is full if we've got more links to send than will
* fit in one chunk
*/
if (links_to_send > 0 && chunk_limit <= links_to_send) {
chunk_full = true;
}
}
return chunk_full;
}
/**
* Goes through any new linked attributes and checks that the target object
* will be known to the client, i.e. we've already sent it in an replication
* chunk. If not, then it adds the target object to the current replication
* chunk. This is only done when the client specifies DRS_GET_TGT.
*/
static WERROR getncchanges_chunk_add_la_targets(struct getncchanges_repl_chunk *repl_chunk,
struct drsuapi_getncchanges_state *getnc_state,
uint32_t start_la_index,
TALLOC_CTX *mem_ctx,
struct ldb_context *sam_ctx,
struct dsdb_schema *schema,
DATA_BLOB *session_key,
struct drsuapi_DsGetNCChangesRequest10 *req10,
uint32_t *local_pas,
struct ldb_dn *machine_dn)
{
int ret;
uint32_t i;
uint32_t max_la_index;
uint32_t max_links;
uint32_t target_count = 0;
WERROR werr = WERR_OK;
static const char * const msg_attrs[] = {
"*",
"nTSecurityDescriptor",
"parentGUID",
"replPropertyMetaData",
DSDB_SECRET_ATTRIBUTES,
NULL };
/*
* A object can potentially link to thousands of targets. Only bother
* checking as many targets as will fit into the current response
*/
max_links = getncchanges_chunk_max_links(repl_chunk);
max_la_index = MIN(getnc_state->la_count,
start_la_index + max_links);
/* loop through any linked attributes to check */
for (i = start_la_index;
(i < max_la_index &&
!getncchanges_chunk_is_full(repl_chunk, getnc_state));
i++) {
struct GUID target_guid;
struct drsuapi_DsReplicaObjectListItemEx *new_objs = NULL;
const struct drsuapi_DsReplicaLinkedAttribute *la;
struct ldb_result *msg_res;
struct ldb_dn *search_dn;
TALLOC_CTX *tmp_ctx;
struct dsdb_dn *dn;
const struct dsdb_attribute *schema_attrib;
NTSTATUS status;
bool same_nc;
la = &getnc_state->la_list[i];
tmp_ctx = talloc_new(mem_ctx);
/*
* Track what linked attribute targets we've checked. We might
* not have time to check them all, so we should only send back
* the ones we've actually checked.
*/
repl_chunk->tgt_la_count = i + 1;
/* get the GUID of the linked attribute's target object */
schema_attrib = dsdb_attribute_by_attributeID_id(schema,
la->attid);
werr = dsdb_dn_la_from_blob(sam_ctx, schema_attrib, schema,
tmp_ctx, la->value.blob, &dn);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__ ": Bad la blob\n"));
return werr;
}
status = dsdb_get_extended_dn_guid(dn->dn, &target_guid, "GUID");
if (!NT_STATUS_IS_OK(status)) {
return ntstatus_to_werror(status);
}
/*
* if the target isn't in the cache, then the client
* might not know about it, so send the target now
*/
werr = dcesrv_drsuapi_obj_cache_exists(getnc_state->obj_cache,
&target_guid);
if (W_ERROR_EQUAL(werr, WERR_OBJECT_NAME_EXISTS)) {
/* target already sent, nothing to do */
TALLOC_FREE(tmp_ctx);
continue;
}
same_nc = dsdb_objects_have_same_nc(sam_ctx, tmp_ctx, dn->dn,
getnc_state->ncRoot_dn);
/* don't try to fetch target objects from another partition */
if (!same_nc) {
TALLOC_FREE(tmp_ctx);
continue;
}
search_dn = ldb_dn_new_fmt(tmp_ctx, sam_ctx, "<GUID=%s>",
GUID_string(tmp_ctx, &target_guid));
W_ERROR_HAVE_NO_MEMORY(search_dn);
ret = drsuapi_search_with_extended_dn(sam_ctx, tmp_ctx,
&msg_res, search_dn,
LDB_SCOPE_BASE,
msg_attrs, NULL);
/*
* Don't fail the replication if we can't find the target.
* This could happen for a one-way linked attribute, if the
* target is deleted and then later expunged (thus, the source
* object can be left with a hanging link). Continue to send
* the the link (the client-side has already tried once with
* GET_TGT, so it should just end up ignoring it).
*/
if (ret == LDB_ERR_NO_SUCH_OBJECT) {
DBG_WARNING("Encountered unknown link target DN %s\n",
ldb_dn_get_extended_linearized(tmp_ctx, dn->dn, 1));
TALLOC_FREE(tmp_ctx);
continue;
} else if (ret != LDB_SUCCESS) {
DBG_ERR("Failed to fetch link target DN %s - %s\n",
ldb_dn_get_extended_linearized(tmp_ctx, dn->dn, 1),
ldb_errstring(sam_ctx));
return WERR_DS_DRA_INCONSISTENT_DIT;
}
/*
* Construct an object, ready to send (this will include
* the object's ancestors as well, if GET_ANC is set)
*/
werr = getncchanges_get_obj_to_send(msg_res->msgs[0], mem_ctx,
sam_ctx, getnc_state,
schema, session_key, req10,
false, local_pas,
machine_dn, &target_guid,
&new_objs);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
if (new_objs != NULL) {
target_count++;
getncchanges_chunk_add_objects(repl_chunk, new_objs);
}
TALLOC_FREE(tmp_ctx);
}
if (target_count > 0) {
DEBUG(3, ("GET_TGT: checked %u link-attrs, added %u target objs\n",
i - start_la_index, target_count));
}
return WERR_OK;
}
/**
* Creates a helper struct used for building a chunk of replication data,
* i.e. used over a single call to dcesrv_drsuapi_DsGetNCChanges().
*/
static struct getncchanges_repl_chunk * getncchanges_chunk_new(TALLOC_CTX *mem_ctx,
struct dcesrv_call_state *dce_call,
struct drsuapi_DsGetNCChangesRequest10 *req10)
{
struct getncchanges_repl_chunk *repl_chunk;
repl_chunk = talloc_zero(mem_ctx, struct getncchanges_repl_chunk);
repl_chunk->start = time(NULL);
repl_chunk->max_objects = lpcfg_parm_int(dce_call->conn->dce_ctx->lp_ctx, NULL,
"drs", "max object sync",
DEFAULT_MAX_OBJECTS);
/*
* The client control here only applies in normal replication, not extended
* operations, which return a fixed set, even if the caller
* sets max_object_count == 0
*/
if (req10->extended_op == DRSUAPI_EXOP_NONE) {
/*
* use this to force single objects at a time, which is useful
* for working out what object is giving problems
*/
if (req10->max_object_count < repl_chunk->max_objects) {
repl_chunk->max_objects = req10->max_object_count;
}
}
repl_chunk->max_links =
lpcfg_parm_int(dce_call->conn->dce_ctx->lp_ctx, NULL,
"drs", "max link sync",
DEFAULT_MAX_LINKS);
repl_chunk->immediate_link_sync =
lpcfg_parm_bool(dce_call->conn->dce_ctx->lp_ctx, NULL,
"drs", "immediate link sync", false);
/*
* Maximum time that we can spend in a getncchanges
* in order to avoid timeout of the other part.
* 10 seconds by default.
*/
repl_chunk->max_wait = lpcfg_parm_int(dce_call->conn->dce_ctx->lp_ctx,
NULL, "drs", "max work time", 10);
return repl_chunk;
}
/*
drsuapi_DsGetNCChanges
see MS-DRSR 4.1.10.5.2 for basic logic of this function
*/
WERROR dcesrv_drsuapi_DsGetNCChanges(struct dcesrv_call_state *dce_call, TALLOC_CTX *mem_ctx,
struct drsuapi_DsGetNCChanges *r)
{
struct auth_session_info *session_info =
dcesrv_call_session_info(dce_call);
struct imessaging_context *imsg_ctx =
dcesrv_imessaging_context(dce_call->conn);
struct drsuapi_DsReplicaObjectIdentifier *ncRoot;
int ret;
uint32_t i, k;
struct dsdb_schema *schema;
struct drsuapi_DsReplicaOIDMapping_Ctr *ctr;
struct getncchanges_repl_chunk *repl_chunk;
NTSTATUS status;
DATA_BLOB session_key;
WERROR werr;
struct dcesrv_handle *h;
struct drsuapi_bind_state *b_state;
struct drsuapi_getncchanges_state *getnc_state;
struct drsuapi_DsGetNCChangesRequest10 *req10;
uint32_t options;
uint32_t link_count = 0;
struct ldb_dn *search_dn = NULL;
bool am_rodc;
enum security_user_level security_level;
struct ldb_context *sam_ctx;
struct dom_sid *user_sid;
bool is_secret_request;
bool is_gc_pas_request;
struct drsuapi_changed_objects *changes;
bool has_get_all_changes = false;
struct GUID invocation_id;
static const struct drsuapi_DsReplicaLinkedAttribute no_linked_attr;
struct dsdb_schema_prefixmap *pfm_remote = NULL;
bool full = true;
uint32_t *local_pas = NULL;
struct ldb_dn *machine_dn = NULL; /* Only used for REPL SECRET EXOP */
DCESRV_PULL_HANDLE_WERR(h, r->in.bind_handle, DRSUAPI_BIND_HANDLE);
b_state = h->data;
/* sam_ctx_system is not present for non-administrator users */
sam_ctx = b_state->sam_ctx_system?b_state->sam_ctx_system:b_state->sam_ctx;
invocation_id = *(samdb_ntds_invocation_id(sam_ctx));
*r->out.level_out = 6;
r->out.ctr->ctr6.linked_attributes_count = 0;
r->out.ctr->ctr6.linked_attributes = discard_const_p(struct drsuapi_DsReplicaLinkedAttribute, &no_linked_attr);
r->out.ctr->ctr6.object_count = 0;
r->out.ctr->ctr6.nc_object_count = 0;
r->out.ctr->ctr6.more_data = false;
r->out.ctr->ctr6.uptodateness_vector = NULL;
r->out.ctr->ctr6.source_dsa_guid = *(samdb_ntds_objectGUID(sam_ctx));
r->out.ctr->ctr6.source_dsa_invocation_id = *(samdb_ntds_invocation_id(sam_ctx));
r->out.ctr->ctr6.first_object = NULL;
/* Check request revision.
*/
switch (r->in.level) {
case 8:
req10 = getncchanges_map_req8(mem_ctx, &r->in.req->req8);
if (req10 == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
break;
case 10:
req10 = &r->in.req->req10;
break;
default:
DEBUG(0,(__location__ ": Request for DsGetNCChanges with unsupported level %u\n",
r->in.level));
return WERR_REVISION_MISMATCH;
}
repl_chunk = getncchanges_chunk_new(mem_ctx, dce_call, req10);
if (repl_chunk == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
/* a RODC doesn't allow for any replication */
ret = samdb_rodc(sam_ctx, &am_rodc);
if (ret == LDB_SUCCESS && am_rodc) {
DEBUG(0,(__location__ ": DsGetNCChanges attempt on RODC\n"));
return WERR_DS_DRA_SOURCE_DISABLED;
}
/* Perform access checks. */
/* TODO: we need to support a sync on a specific non-root
* DN. We'll need to find the real partition root here */
ncRoot = req10->naming_context;
if (ncRoot == NULL) {
DEBUG(0,(__location__ ": Request for DsGetNCChanges with no NC\n"));
return WERR_DS_DRA_INVALID_PARAMETER;
}
if (samdb_ntds_options(sam_ctx, &options) != LDB_SUCCESS) {
return WERR_DS_DRA_INTERNAL_ERROR;
}
if ((options & DS_NTDSDSA_OPT_DISABLE_OUTBOUND_REPL) &&
!(req10->replica_flags & DRSUAPI_DRS_SYNC_FORCED)) {
return WERR_DS_DRA_SOURCE_DISABLED;
}
user_sid = &session_info->security_token->sids[PRIMARY_USER_SID_INDEX];
/* all clients must have GUID_DRS_GET_CHANGES */
werr = drs_security_access_check_nc_root(sam_ctx,
mem_ctx,
session_info->security_token,
req10->naming_context,
GUID_DRS_GET_CHANGES);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
if (dsdb_functional_level(sam_ctx) >= DS_DOMAIN_FUNCTION_2008) {
full = req10->partial_attribute_set == NULL &&
req10->partial_attribute_set_ex == NULL;
} else {
full = (options & DRSUAPI_DRS_WRIT_REP) != 0;
}
werr = dsdb_schema_pfm_from_drsuapi_pfm(&req10->mapping_ctr, true,
mem_ctx, &pfm_remote, NULL);
/* We were supplied a partial attribute set, without the prefix map! */
if (!full && !W_ERROR_IS_OK(werr)) {
if (req10->mapping_ctr.num_mappings == 0) {
/*
* Despite the fact MS-DRSR specifies that this shouldn't
* happen, Windows RODCs will in fact not provide a prefixMap.
*/
DEBUG(5,(__location__ ": Failed to provide a remote prefixMap,"
" falling back to local prefixMap\n"));
} else {
DEBUG(0,(__location__ ": Failed to decode remote prefixMap: %s\n",
win_errstr(werr)));
return werr;
}
}
/* allowed if the GC PAS and client has
GUID_DRS_GET_FILTERED_ATTRIBUTES */
werr = dcesrv_drsuapi_is_gc_pas_request(b_state, req10, pfm_remote, &is_gc_pas_request);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
if (is_gc_pas_request) {
werr = drs_security_access_check_nc_root(sam_ctx,
mem_ctx,
session_info->security_token,
req10->naming_context,
GUID_DRS_GET_FILTERED_ATTRIBUTES);
if (W_ERROR_IS_OK(werr)) {
goto allowed;
}
}
werr = dcesrv_drsuapi_is_reveal_secrets_request(b_state, req10,
pfm_remote,
&is_secret_request);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
if (is_secret_request) {
werr = drs_security_access_check_nc_root(sam_ctx,
mem_ctx,
session_info->security_token,
req10->naming_context,
GUID_DRS_GET_ALL_CHANGES);
if (!W_ERROR_IS_OK(werr)) {
/* Only bail if this is not a EXOP_REPL_SECRET */
if (req10->extended_op != DRSUAPI_EXOP_REPL_SECRET) {
return werr;
}
} else {
has_get_all_changes = true;
}
}
allowed:
/* for non-administrator replications, check that they have
given the correct source_dsa_invocation_id */
security_level = security_session_user_level(session_info,
samdb_domain_sid(sam_ctx));
if (security_level == SECURITY_RO_DOMAIN_CONTROLLER) {
if (req10->replica_flags & DRSUAPI_DRS_WRIT_REP) {
/* we rely on this flag being unset for RODC requests */
req10->replica_flags &= ~DRSUAPI_DRS_WRIT_REP;
}
}
if (req10->replica_flags & DRSUAPI_DRS_FULL_SYNC_PACKET) {
/* Ignore the _in_ uptpdateness vector*/
req10->uptodateness_vector = NULL;
}
if (GUID_all_zero(&req10->source_dsa_invocation_id)) {
req10->source_dsa_invocation_id = invocation_id;
}
if (!GUID_equal(&req10->source_dsa_invocation_id, &invocation_id)) {
/*
* The given highwatermark is only valid relative to the
* specified source_dsa_invocation_id.
*/
ZERO_STRUCT(req10->highwatermark);
}
getnc_state = b_state->getncchanges_state;
/* see if a previous replication has been abandoned */
if (getnc_state) {
struct ldb_dn *new_dn = drs_ObjectIdentifier_to_dn(getnc_state, sam_ctx, ncRoot);
if (ldb_dn_compare(new_dn, getnc_state->ncRoot_dn) != 0) {
DEBUG(0,(__location__ ": DsGetNCChanges 2nd replication on different DN %s %s (last_dn %s)\n",
ldb_dn_get_linearized(new_dn),
ldb_dn_get_linearized(getnc_state->ncRoot_dn),
ldb_dn_get_linearized(getnc_state->last_dn)));
TALLOC_FREE(getnc_state);
b_state->getncchanges_state = NULL;
}
}
if (getnc_state) {
ret = drsuapi_DsReplicaHighWaterMark_cmp(&getnc_state->last_hwm,
&req10->highwatermark);
if (ret != 0) {
DEBUG(0,(__location__ ": DsGetNCChanges 2nd replication "
"on DN %s %s highwatermark (last_dn %s)\n",
ldb_dn_get_linearized(getnc_state->ncRoot_dn),
(ret > 0) ? "older" : "newer",
ldb_dn_get_linearized(getnc_state->last_dn)));
TALLOC_FREE(getnc_state);
b_state->getncchanges_state = NULL;
}
}
if (getnc_state == NULL) {
struct ldb_result *res = NULL;
const char *attrs[] = {
"instanceType",
"objectGuID",
NULL
};
uint32_t nc_instanceType;
struct ldb_dn *ncRoot_dn;
ncRoot_dn = drs_ObjectIdentifier_to_dn(mem_ctx, sam_ctx, ncRoot);
if (ncRoot_dn == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
ret = dsdb_search_dn(sam_ctx, mem_ctx, &res,
ncRoot_dn, attrs,
DSDB_SEARCH_SHOW_DELETED |
DSDB_SEARCH_SHOW_RECYCLED);
if (ret != LDB_SUCCESS) {
DBG_WARNING("Failed to find ncRoot_dn %s\n",
ldb_dn_get_linearized(ncRoot_dn));
return WERR_DS_DRA_BAD_DN;
}
nc_instanceType = ldb_msg_find_attr_as_int(res->msgs[0],
"instanceType",
0);
if (req10->extended_op != DRSUAPI_EXOP_NONE) {
r->out.ctr->ctr6.extended_ret = DRSUAPI_EXOP_ERR_SUCCESS;
}
/*
* This is the first replication cycle and it is
* a good place to handle extended operations
*
* FIXME: we don't fully support extended operations yet
*/
switch (req10->extended_op) {
case DRSUAPI_EXOP_NONE:
if ((nc_instanceType & INSTANCE_TYPE_IS_NC_HEAD) == 0) {
const char *dn_str
= ldb_dn_get_linearized(ncRoot_dn);
DBG_NOTICE("Rejecting full replication on "
"not NC %s", dn_str);
return WERR_DS_CANT_FIND_EXPECTED_NC;
}
break;
case DRSUAPI_EXOP_FSMO_RID_ALLOC:
werr = getncchanges_rid_alloc(b_state, mem_ctx, req10, &r->out.ctr->ctr6, &search_dn);
W_ERROR_NOT_OK_RETURN(werr);
if (r->out.ctr->ctr6.extended_ret != DRSUAPI_EXOP_ERR_SUCCESS) {
return WERR_OK;
}
break;
case DRSUAPI_EXOP_REPL_SECRET:
werr = getncchanges_repl_secret(b_state, mem_ctx, req10,
user_sid,
&r->out.ctr->ctr6,
has_get_all_changes,
&machine_dn);
r->out.result = werr;
W_ERROR_NOT_OK_RETURN(werr);
break;
case DRSUAPI_EXOP_FSMO_REQ_ROLE:
werr = getncchanges_change_master(b_state, mem_ctx, req10, &r->out.ctr->ctr6);
W_ERROR_NOT_OK_RETURN(werr);
if (r->out.ctr->ctr6.extended_ret != DRSUAPI_EXOP_ERR_SUCCESS) {
return WERR_OK;
}
break;
case DRSUAPI_EXOP_FSMO_RID_REQ_ROLE:
werr = getncchanges_change_master(b_state, mem_ctx, req10, &r->out.ctr->ctr6);
W_ERROR_NOT_OK_RETURN(werr);
if (r->out.ctr->ctr6.extended_ret != DRSUAPI_EXOP_ERR_SUCCESS) {
return WERR_OK;
}
break;
case DRSUAPI_EXOP_FSMO_REQ_PDC:
werr = getncchanges_change_master(b_state, mem_ctx, req10, &r->out.ctr->ctr6);
W_ERROR_NOT_OK_RETURN(werr);
if (r->out.ctr->ctr6.extended_ret != DRSUAPI_EXOP_ERR_SUCCESS) {
return WERR_OK;
}
break;
case DRSUAPI_EXOP_REPL_OBJ:
werr = getncchanges_repl_obj(b_state, mem_ctx, req10, user_sid, &r->out.ctr->ctr6);
r->out.result = werr;
W_ERROR_NOT_OK_RETURN(werr);
break;
case DRSUAPI_EXOP_FSMO_ABANDON_ROLE:
DEBUG(0,(__location__ ": Request for DsGetNCChanges unsupported extended op 0x%x\n",
(unsigned)req10->extended_op));
return WERR_DS_DRA_NOT_SUPPORTED;
}
/* Initialize the state we'll store over the replication cycle */
getnc_state = talloc_zero(b_state, struct drsuapi_getncchanges_state);
if (getnc_state == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
b_state->getncchanges_state = getnc_state;
getnc_state->ncRoot_dn = ncRoot_dn;
talloc_steal(getnc_state, ncRoot_dn);
getnc_state->ncRoot_guid = samdb_result_guid(res->msgs[0],
"objectGUID");
ncRoot->guid = getnc_state->ncRoot_guid;
/* find out if we are to replicate Schema NC */
ret = ldb_dn_compare_base(ldb_get_schema_basedn(sam_ctx),
ncRoot_dn);
getnc_state->is_schema_nc = (0 == ret);
TALLOC_FREE(res);
}
if (!ldb_dn_validate(getnc_state->ncRoot_dn) ||
ldb_dn_is_null(getnc_state->ncRoot_dn)) {
DEBUG(0,(__location__ ": Bad DN '%s'\n",
drs_ObjectIdentifier_to_string(mem_ctx, ncRoot)));
return WERR_DS_DRA_INVALID_PARAMETER;
}
ncRoot->guid = getnc_state->ncRoot_guid;
/* we need the session key for encrypting password attributes */
status = dcesrv_auth_session_key(dce_call, &session_key);
if (!NT_STATUS_IS_OK(status)) {
DEBUG(0,(__location__ ": Failed to get session key\n"));
return WERR_DS_DRA_INTERNAL_ERROR;
}
/*
TODO: MS-DRSR section 4.1.10.1.1
Work out if this is the start of a new cycle */
if (getnc_state->guids == NULL) {
const char *extra_filter;
struct ldb_result *search_res = NULL;
static const struct drsuapi_DsReplicaCursorCtrEx empty_udv;
const struct drsuapi_DsReplicaCursorCtrEx *udv = NULL;
extra_filter = lpcfg_parm_string(dce_call->conn->dce_ctx->lp_ctx, NULL, "drs", "object filter");
if (req10->extended_op == DRSUAPI_EXOP_NONE) {
if (req10->uptodateness_vector != NULL) {
udv = req10->uptodateness_vector;
} else {
udv = &empty_udv;
}
getnc_state->min_usn = req10->highwatermark.highest_usn;
for (i = 0; i < udv->count; i++) {
bool match;
const struct drsuapi_DsReplicaCursor *cur =
&udv->cursors[i];
match = GUID_equal(&invocation_id,
&cur->source_dsa_invocation_id);
if (!match) {
continue;
}
if (cur->highest_usn > getnc_state->min_usn) {
getnc_state->min_usn = cur->highest_usn;
}
break;
}
} else {
/* We do not want REPL_SECRETS or REPL_SINGLE to return empty-handed */
udv = &empty_udv;
getnc_state->min_usn = 0;
}
getnc_state->max_usn = getnc_state->min_usn;
getnc_state->final_udv = talloc_zero(getnc_state,
struct drsuapi_DsReplicaCursor2CtrEx);
if (getnc_state->final_udv == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
werr = get_nc_changes_udv(sam_ctx, getnc_state->ncRoot_dn,
getnc_state->final_udv);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
if (req10->extended_op == DRSUAPI_EXOP_NONE) {
werr = getncchanges_collect_objects(b_state, mem_ctx, req10,
search_dn, extra_filter,
&search_res);
} else {
werr = getncchanges_collect_objects_exop(b_state, mem_ctx, req10,
&r->out.ctr->ctr6,
search_dn, extra_filter,
&search_res);
}
W_ERROR_NOT_OK_RETURN(werr);
/* extract out the GUIDs list */
getnc_state->num_records = search_res ? search_res->count : 0;
getnc_state->guids = talloc_array(getnc_state, struct GUID, getnc_state->num_records);
W_ERROR_HAVE_NO_MEMORY(getnc_state->guids);
changes = talloc_array(getnc_state,
struct drsuapi_changed_objects,
getnc_state->num_records);
W_ERROR_HAVE_NO_MEMORY(changes);
for (i=0; i<getnc_state->num_records; i++) {
changes[i].dn = search_res->msgs[i]->dn;
changes[i].guid = samdb_result_guid(search_res->msgs[i], "objectGUID");
changes[i].usn = ldb_msg_find_attr_as_uint64(search_res->msgs[i], "uSNChanged", 0);
if (changes[i].usn > getnc_state->max_usn) {
getnc_state->max_usn = changes[i].usn;
}
}
/* RID_ALLOC returns 3 objects in a fixed order */
if (req10->extended_op == DRSUAPI_EXOP_FSMO_RID_ALLOC) {
/* Do nothing */
} else {
LDB_TYPESAFE_QSORT(changes,
getnc_state->num_records,
getnc_state,
site_res_cmp_usn_order);
}
for (i=0; i < getnc_state->num_records; i++) {
getnc_state->guids[i] = changes[i].guid;
if (GUID_all_zero(&getnc_state->guids[i])) {
DEBUG(2,("getncchanges: bad objectGUID from %s\n",
ldb_dn_get_linearized(search_res->msgs[i]->dn)));
return WERR_DS_DRA_INTERNAL_ERROR;
}
}
getnc_state->final_hwm.tmp_highest_usn = getnc_state->max_usn;
getnc_state->final_hwm.reserved_usn = 0;
getnc_state->final_hwm.highest_usn = getnc_state->max_usn;
talloc_free(search_res);
talloc_free(changes);
if (req10->extended_op == DRSUAPI_EXOP_NONE) {
getnc_state->is_get_anc =
((req10->replica_flags & DRSUAPI_DRS_GET_ANC) != 0);
getnc_state->is_get_tgt =
((req10->more_flags & DRSUAPI_DRS_GET_TGT) != 0);
}
/*
* when using GET_ANC or GET_TGT, cache the objects that have
* been already sent, to avoid sending them multiple times
*/
if (getnc_state->is_get_anc || getnc_state->is_get_tgt) {
DEBUG(3,("Using object cache, GET_ANC %u, GET_TGT %u\n",
getnc_state->is_get_anc,
getnc_state->is_get_tgt));
getnc_state->obj_cache = db_open_rbt(getnc_state);
if (getnc_state->obj_cache == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
}
}
if (req10->uptodateness_vector) {
/* make sure its sorted */
TYPESAFE_QSORT(req10->uptodateness_vector->cursors,
req10->uptodateness_vector->count,
drsuapi_DsReplicaCursor_compare);
}
/* Prefix mapping */
schema = dsdb_get_schema(sam_ctx, mem_ctx);
if (!schema) {
DEBUG(0,("No schema in sam_ctx\n"));
return WERR_DS_DRA_INTERNAL_ERROR;
}
r->out.ctr->ctr6.naming_context = talloc(mem_ctx, struct drsuapi_DsReplicaObjectIdentifier);
if (r->out.ctr->ctr6.naming_context == NULL) {
return WERR_NOT_ENOUGH_MEMORY;
}
*r->out.ctr->ctr6.naming_context = *ncRoot;
/* find the SID if there is one */
dsdb_find_sid_by_dn(sam_ctx, getnc_state->ncRoot_dn, &r->out.ctr->ctr6.naming_context->sid);
dsdb_get_oid_mappings_drsuapi(schema, true, mem_ctx, &ctr);
r->out.ctr->ctr6.mapping_ctr = *ctr;
r->out.ctr->ctr6.source_dsa_guid = *(samdb_ntds_objectGUID(sam_ctx));
r->out.ctr->ctr6.source_dsa_invocation_id = *(samdb_ntds_invocation_id(sam_ctx));
r->out.ctr->ctr6.old_highwatermark = req10->highwatermark;
r->out.ctr->ctr6.new_highwatermark = req10->highwatermark;
/*
* If the client has already set GET_TGT then we know they can handle
* receiving the linked attributes interleaved with the source objects
*/
if (getnc_state->is_get_tgt) {
repl_chunk->immediate_link_sync = true;
}
if (req10->partial_attribute_set != NULL) {
struct dsdb_syntax_ctx syntax_ctx;
uint32_t j = 0;
dsdb_syntax_ctx_init(&syntax_ctx, sam_ctx, schema);
syntax_ctx.pfm_remote = pfm_remote;
local_pas = talloc_array(b_state, uint32_t, req10->partial_attribute_set->num_attids);
for (j = 0; j < req10->partial_attribute_set->num_attids; j++) {
getncchanges_attid_remote_to_local(schema,
&syntax_ctx,
req10->partial_attribute_set->attids[j],
(enum drsuapi_DsAttributeId *)&local_pas[j],
NULL);
}
TYPESAFE_QSORT(local_pas,
req10->partial_attribute_set->num_attids,
uint32_t_ptr_cmp);
}
/*
* Check in case we're still processing the links from an object in the
* previous chunk. We want to send the links (and any targets needed)
* before moving on to the next object.
*/
if (getnc_state->is_get_tgt) {
werr = getncchanges_chunk_add_la_targets(repl_chunk,
getnc_state,
getnc_state->la_idx,
mem_ctx, sam_ctx,
schema, &session_key,
req10, local_pas,
machine_dn);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
for (i=getnc_state->num_processed;
i<getnc_state->num_records &&
!getncchanges_chunk_is_full(repl_chunk, getnc_state);
i++) {
struct drsuapi_DsReplicaObjectListItemEx *new_objs = NULL;
struct ldb_message *msg;
static const char * const msg_attrs[] = {
"*",
"nTSecurityDescriptor",
"parentGUID",
"replPropertyMetaData",
DSDB_SECRET_ATTRIBUTES,
NULL };
struct ldb_result *msg_res;
struct ldb_dn *msg_dn;
bool obj_already_sent = false;
TALLOC_CTX *tmp_ctx = talloc_new(mem_ctx);
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
uint32_t old_la_index;
msg_dn = ldb_dn_new_fmt(tmp_ctx, sam_ctx, "<GUID=%s>",
GUID_string(tmp_ctx, &getnc_state->guids[i]));
W_ERROR_HAVE_NO_MEMORY(msg_dn);
/*
* by re-searching here we avoid having a lot of full
* records in memory between calls to getncchanges.
*
* We expect that we may get some objects that vanish
* (tombstone expunge) between the first and second
* check.
*/
ret = drsuapi_search_with_extended_dn(sam_ctx, tmp_ctx, &msg_res,
msg_dn,
LDB_SCOPE_BASE, msg_attrs, NULL);
if (ret != LDB_SUCCESS) {
if (ret != LDB_ERR_NO_SUCH_OBJECT) {
DEBUG(1,("getncchanges: failed to fetch DN %s - %s\n",
ldb_dn_get_extended_linearized(tmp_ctx, msg_dn, 1),
ldb_errstring(sam_ctx)));
}
TALLOC_FREE(tmp_ctx);
continue;
}
if (msg_res->count == 0) {
DEBUG(1,("getncchanges: got LDB_SUCCESS but failed"
"to get any results in fetch of DN "
"%s (race with tombstone expunge?)\n",
ldb_dn_get_extended_linearized(tmp_ctx,
msg_dn, 1)));
TALLOC_FREE(tmp_ctx);
continue;
}
msg = msg_res->msgs[0];
/*
* Check if we've already sent the object as an ancestor of
* another object. If so, we don't need to send it again
*/
if (getnc_state->obj_cache != NULL) {
werr = dcesrv_drsuapi_obj_cache_exists(getnc_state->obj_cache,
&getnc_state->guids[i]);
if (W_ERROR_EQUAL(werr, WERR_OBJECT_NAME_EXISTS)) {
obj_already_sent = true;
}
}
if (!obj_already_sent) {
bool max_wait_reached;
max_wait_reached = getncchanges_chunk_timed_out(repl_chunk);
/*
* Construct an object, ready to send (this will include
* the object's ancestors as well, if needed)
*/
werr = getncchanges_get_obj_to_send(msg, mem_ctx, sam_ctx,
getnc_state, schema,
&session_key, req10,
max_wait_reached,
local_pas, machine_dn,
&getnc_state->guids[i],
&new_objs);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
2009-09-10 13:41:48 +04:00
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
old_la_index = getnc_state->la_count;
/*
* We've reached the USN where this object naturally occurs.
* Regardless of whether we've already sent the object (as an
* ancestor), we add its links and update the HWM at this point
*/
werr = get_nc_changes_add_links(sam_ctx, getnc_state,
getnc_state->is_schema_nc,
schema, getnc_state->min_usn,
req10->replica_flags,
msg,
&getnc_state->la_list,
&getnc_state->la_count,
req10->uptodateness_vector);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
dcesrv_drsuapi_update_highwatermark(msg,
getnc_state->max_usn,
&r->out.ctr->ctr6.new_highwatermark);
if (new_objs != NULL) {
2009-09-10 13:41:48 +04:00
/*
* Add the object (and, if GET_ANC, any parents it may
* have) into the current chunk of replication data
*/
getncchanges_chunk_add_objects(repl_chunk, new_objs);
talloc_free(getnc_state->last_dn);
getnc_state->last_dn = talloc_move(getnc_state, &msg->dn);
}
DEBUG(8,(__location__ ": %s object %s\n",
new_objs ? "replicating" : "skipping send of",
ldb_dn_get_linearized(msg->dn)));
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
getnc_state->total_links += (getnc_state->la_count - old_la_index);
/*
* If the GET_TGT flag was set, check any new links added to
* make sure the client knows about the link target object
*/
if (getnc_state->is_get_tgt) {
werr = getncchanges_chunk_add_la_targets(repl_chunk,
getnc_state,
old_la_index,
mem_ctx, sam_ctx,
schema, &session_key,
req10, local_pas,
machine_dn);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
}
TALLOC_FREE(tmp_ctx);
}
/* copy the constructed object list into the response message */
r->out.ctr->ctr6.object_count = repl_chunk->object_count;
r->out.ctr->ctr6.first_object = repl_chunk->object_list;
getnc_state->num_processed = i;
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
if (i < getnc_state->num_records) {
r->out.ctr->ctr6.more_data = true;
}
/* the client can us to call UpdateRefs on its behalf to
re-establish monitoring of the NC */
if ((req10->replica_flags & (DRSUAPI_DRS_ADD_REF | DRSUAPI_DRS_REF_GCSPN)) &&
!GUID_all_zero(&req10->destination_dsa_guid)) {
struct drsuapi_DsReplicaUpdateRefsRequest1 ureq;
DEBUG(3,("UpdateRefs on getncchanges for %s\n",
GUID_string(mem_ctx, &req10->destination_dsa_guid)));
ureq.naming_context = ncRoot;
ureq.dest_dsa_dns_name = samdb_ntds_msdcs_dns_name(sam_ctx, mem_ctx,
&req10->destination_dsa_guid);
if (!ureq.dest_dsa_dns_name) {
return WERR_NOT_ENOUGH_MEMORY;
}
ureq.dest_dsa_guid = req10->destination_dsa_guid;
ureq.options = DRSUAPI_DRS_ADD_REF |
DRSUAPI_DRS_ASYNC_OP |
DRSUAPI_DRS_GETCHG_CHECK;
/* we also need to pass through the
DRSUAPI_DRS_REF_GCSPN bit so that repsTo gets flagged
to send notifies using the GC SPN */
ureq.options |= (req10->replica_flags & DRSUAPI_DRS_REF_GCSPN);
werr = drsuapi_UpdateRefs(imsg_ctx,
dce_call->event_ctx,
b_state,
mem_ctx,
&ureq);
if (!W_ERROR_IS_OK(werr)) {
DEBUG(0,(__location__ ": Failed UpdateRefs on %s for %s in DsGetNCChanges - %s\n",
drs_ObjectIdentifier_to_string(mem_ctx, ncRoot), ureq.dest_dsa_dns_name,
win_errstr(werr)));
}
}
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
/*
* Work out how many links we can send in this chunk. The default is to
* send all the links last, but there is a config option to send them
* immediately, in the same chunk as their source object
*/
if (!r->out.ctr->ctr6.more_data || repl_chunk->immediate_link_sync) {
link_count = getncchanges_chunk_links_pending(repl_chunk,
getnc_state);
link_count = MIN(link_count,
getncchanges_chunk_max_links(repl_chunk));
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
}
/* If we've got linked attributes to send, add them now */
if (link_count > 0) {
struct la_for_sorting *la_sorted;
/*
* Grab a chunk of linked attributes off the list and put them
* in sorted array, ready to send
*/
werr = getncchanges_get_sorted_array(&getnc_state->la_list[getnc_state->la_idx],
link_count,
sam_ctx, getnc_state,
schema,
&la_sorted);
if (!W_ERROR_IS_OK(werr)) {
return werr;
}
r->out.ctr->ctr6.linked_attributes_count = link_count;
r->out.ctr->ctr6.linked_attributes = talloc_array(r->out.ctr, struct drsuapi_DsReplicaLinkedAttribute, link_count);
if (r->out.ctr->ctr6.linked_attributes == NULL) {
DEBUG(0, ("Out of memory allocating %u linked attributes for output", link_count));
return WERR_NOT_ENOUGH_MEMORY;
}
for (k = 0; k < link_count; k++) {
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
r->out.ctr->ctr6.linked_attributes[k] = *la_sorted[k].link;
}
getnc_state->la_idx += link_count;
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
getnc_state->links_given += link_count;
if (getnc_state->la_idx < getnc_state->la_count) {
r->out.ctr->ctr6.more_data = true;
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
} else {
/*
* We've now sent all the links seen so far, so we can
* reset la_list back to an empty list again. Note that
* the steal means the linked attribute memory gets
* freed after this RPC message is sent on the wire.
*/
talloc_steal(mem_ctx, getnc_state->la_list);
getnc_state->la_list = NULL;
getnc_state->la_idx = 0;
getnc_state->la_count = 0;
}
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
TALLOC_FREE(la_sorted);
}
if (req10->replica_flags & DRSUAPI_DRS_GET_NC_SIZE) {
/*
* TODO: This implementation is wrong
* we should find out the total number of
* objects and links in the whole naming context
* at the start of the cycle and return these
* values in each message.
*
* For now we keep our current strategy and return
* the number of objects for this cycle and the number
* of links we found so far during the cycle.
*/
r->out.ctr->ctr6.nc_object_count = getnc_state->num_records;
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
r->out.ctr->ctr6.nc_linked_attributes_count = getnc_state->total_links;
}
if (!r->out.ctr->ctr6.more_data) {
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
/* this is the last response in the replication cycle */
r->out.ctr->ctr6.new_highwatermark = getnc_state->final_hwm;
r->out.ctr->ctr6.uptodateness_vector = talloc_move(mem_ctx,
&getnc_state->final_udv);
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
/*
* Free the state info stored for the replication cycle. Note
* that the RPC message we're sending contains links stored in
* getnc_state. mem_ctx is local to this RPC call, so the memory
* will get freed after the RPC message is sent on the wire.
*/
talloc_steal(mem_ctx, getnc_state);
b_state->getncchanges_state = NULL;
} else {
ret = drsuapi_DsReplicaHighWaterMark_cmp(&r->out.ctr->ctr6.old_highwatermark,
&r->out.ctr->ctr6.new_highwatermark);
if (ret == 0) {
/*
* We need to make sure that we never return the
* same highwatermark within the same replication
* cycle more than once. Otherwise we cannot detect
* when the client uses an unexptected highwatermark.
*
* This is a HACK which is needed because our
* object ordering is wrong and set tmp_highest_usn
* to a value that is higher than what we already
* sent to the client (destination dsa).
*/
r->out.ctr->ctr6.new_highwatermark.reserved_usn += 1;
}
getnc_state->last_hwm = r->out.ctr->ctr6.new_highwatermark;
}
if (req10->extended_op != DRSUAPI_EXOP_NONE) {
r->out.ctr->ctr6.uptodateness_vector = NULL;
r->out.ctr->ctr6.nc_object_count = 0;
ZERO_STRUCT(r->out.ctr->ctr6.new_highwatermark);
}
TALLOC_FREE(repl_chunk);
DEBUG(r->out.ctr->ctr6.more_data?4:2,
("DsGetNCChanges with uSNChanged >= %llu flags 0x%08x on %s gave %u objects (done %u/%u) %u links (done %u/%u (as %s))\n",
(unsigned long long)(req10->highwatermark.highest_usn+1),
req10->replica_flags, drs_ObjectIdentifier_to_string(mem_ctx, ncRoot),
r->out.ctr->ctr6.object_count,
i, r->out.ctr->ctr6.more_data?getnc_state->num_records:i,
r->out.ctr->ctr6.linked_attributes_count,
getncchanges.c: Send linked attributes in each chunk Instead of sending all the linked attributes at the end, add a configurable option to send the links in each replication chunk. The benefits of this approach are: - it can reduce memory overhead, as we don't have to keep all the links in memory over the entire replication cycle. - the client should never end up knowing about objects but not their links. (Although we're not sure that this has actually resulted in replication problems, i.e. missing links). Note that until we support GET_TGT, this approach can mean we now send a link where the client doesn't know about the target object, causing the client to siliently drop that linked attribute. Hence, this option is switched off by default. Implementation-wise, this code works fairly the same as before. Instead of sorting the entire getnc_state->la_sorted array at the end and then splitting it up over chunks, we now split the links up over chunks and then sort them when we copy them into the message. This should be OK, as I believe the MS-DRSR Doc says the links in the message should be sorted (rather than sorting *all* the links overall). Windows behaviour seems to chunk the links based on USN and then sort them. getnc_state->la_idx now tracks which links in getnc_state->la_list[] have already been sent (instead of tracking getnc_state->la_sorted). This means the la_sorted array no longer needs to be stored in getnc_state and we can free the array's memory once we've copied the links into the message. Unfortunately, the link_given/link_total debug no longer reports the correct information, so I've moved these into getncchanges_state struct (and now free the struct a bit later so it's safe to reference in the debug). The vampire_dc testenv has been updated to use this new behaviour. Signed-off-by: Tim Beale <timbeale@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org> Autobuild-Date(master): Fri Sep 15 10:07:33 CEST 2017 on sn-devel-144
2017-06-07 01:46:47 +03:00
getnc_state->links_given, getnc_state->total_links,
dom_sid_string(mem_ctx, user_sid)));
#if 0
if (!r->out.ctr->ctr6.more_data && req10->extended_op != DRSUAPI_EXOP_NONE) {
NDR_PRINT_FUNCTION_DEBUG(drsuapi_DsGetNCChanges, NDR_BOTH, r);
}
#endif
return WERR_OK;
}