1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-21 13:34:40 +03:00
lvm2/tools/pvcreate.c
David Teigland a7c45ddc59 lvmetad: two phase vg_update
Previously, a command sent lvmetad new VG metadata in vg_commit().
In vg_commit(), devices are suspended, so any memory allocation
done by the command while sending to lvmetad, or by lvmetad while
updating its cache could deadlock if memory reclaim was triggered.

Now lvmetad is updated in unlock_vg(), after devices are resumed.
The new method for updating VG metadata in lvmetad is in two phases:

1. In vg_write(), before devices are suspended, the command sends
   lvmetad a short message ("set_vg_info") telling it what the new
   VG seqno will be.  lvmetad sees that the seqno is newer than
   the seqno of its cached VG, so it sets the INVALID flag for the
   cached VG.  If sending the message to lvmetad fails, the command
   fails before the metadata is committed and the change is not made.
   If sending the message succeeds, vg_commit() is called.

2. In unlock_vg(), after devices are resumed, the command sends
   lvmetad the standard vg_update message with the new metadata.
   lvmetad sees that the seqno in the new metadata matches the
   seqno it saved from set_vg_info, and knows it has the latest
   copy, so it clears the INVALID flag for the cached VG.

If a command fails between 1 and 2 (after committing the VG on disk,
but before sending lvmetad the new metadata), the cached VG retains
the INVALID flag in lvmetad.  A subsequent command will read the
cached VG from lvmetad, see the INVALID flag, ignore the cached
copy, read the VG from disk instead, update the lvmetad copy
with the latest copy from disk, (this clears the INVALID flag
in lvmetad), and use the correct VG metadata for the command.

(This INVALID mechanism already existed for use by lvmlockd.)
2016-06-28 02:30:31 +01:00

166 lines
4.5 KiB
C

/*
* Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
* Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
*
* This file is part of LVM2.
*
* This copyrighted material is made available to anyone wishing to use,
* modify, copy, or redistribute it subject to the terms and conditions
* of the GNU Lesser General Public License v.2.1.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "tools.h"
/*
* Intial sanity checking of recovery-related command-line arguments.
* These args are: --restorefile, --uuid, and --physicalvolumesize
*
* Output arguments:
* pp: structure allocated by caller, fields written / validated here
*/
static int pvcreate_restore_params_from_args(struct cmd_context *cmd, int argc,
struct pvcreate_params *pp)
{
pp->restorefile = arg_str_value(cmd, restorefile_ARG, NULL);
if (arg_is_set(cmd, restorefile_ARG) && !arg_is_set(cmd, uuidstr_ARG)) {
log_error("--uuid is required with --restorefile");
return 0;
}
if (!arg_is_set(cmd, restorefile_ARG) && arg_is_set(cmd, uuidstr_ARG)) {
if (!arg_is_set(cmd, norestorefile_ARG) &&
find_config_tree_bool(cmd, devices_require_restorefile_with_uuid_CFG, NULL)) {
log_error("--restorefile is required with --uuid");
return 0;
}
}
if (arg_is_set(cmd, uuidstr_ARG) && argc != 1) {
log_error("Can only set uuid on one volume at once");
return 0;
}
if (arg_is_set(cmd, uuidstr_ARG)) {
pp->uuid_str = arg_str_value(cmd, uuidstr_ARG, "");
if (!id_read_format(&pp->pva.id, pp->uuid_str))
return 0;
pp->pva.idp = &pp->pva.id;
}
if (arg_sign_value(cmd, physicalvolumesize_ARG, SIGN_NONE) == SIGN_MINUS) {
log_error("Physical volume size may not be negative");
return 0;
}
pp->pva.size = arg_uint64_value(cmd, physicalvolumesize_ARG, UINT64_C(0));
if (arg_is_set(cmd, restorefile_ARG) || arg_is_set(cmd, uuidstr_ARG))
pp->zero = 0;
return 1;
}
static int pvcreate_restore_params_from_backup(struct cmd_context *cmd,
struct pvcreate_params *pp)
{
struct volume_group *vg;
struct pv_list *existing_pvl;
/*
* When restoring a PV, params need to be read from a backup file.
*/
if (!pp->restorefile)
return 1;
if (!(vg = backup_read_vg(cmd, NULL, pp->restorefile))) {
log_error("Unable to read volume group from %s", pp->restorefile);
return 0;
}
if (!(existing_pvl = find_pv_in_vg_by_uuid(vg, &pp->pva.id))) {
release_vg(vg);
log_error("Can't find uuid %s in backup file %s",
pp->uuid_str, pp->restorefile);
return 0;
}
pp->pva.ba_start = pv_ba_start(existing_pvl->pv);
pp->pva.ba_size = pv_ba_size(existing_pvl->pv);
pp->pva.pe_start = pv_pe_start(existing_pvl->pv);
pp->pva.extent_size = pv_pe_size(existing_pvl->pv);
pp->pva.extent_count = pv_pe_count(existing_pvl->pv);
release_vg(vg);
return 1;
}
int pvcreate(struct cmd_context *cmd, int argc, char **argv)
{
struct processing_handle *handle;
struct pvcreate_params pp;
int ret;
if (!argc) {
log_error("Please enter a physical volume path.");
return 0;
}
/*
* Device info needs to be available for reading the VG backup file in
* pvcreate_restore_params_from_backup.
*/
lvmcache_seed_infos_from_lvmetad(cmd);
/*
* Five kinds of pvcreate param values:
* 1. defaults
* 2. recovery-related command line args
* 3. recovery-related args from backup file
* 4. normal command line args
* (this also checks some settings from 2 & 3)
* 5. argc/argv free args specifying devices
*/
pvcreate_params_set_defaults(&pp);
if (!pvcreate_restore_params_from_args(cmd, argc, &pp))
return EINVALID_CMD_LINE;
if (!pvcreate_restore_params_from_backup(cmd, &pp))
return EINVALID_CMD_LINE;
if (!pvcreate_params_from_args(cmd, &pp))
return EINVALID_CMD_LINE;
pp.pv_count = argc;
pp.pv_names = argv;
/*
* Needed to change the set of orphan PVs.
* (disable afterward to prevent process_each_pv from doing
* a shared global lock since it's already acquired it ex.)
*/
if (!lockd_gl(cmd, "ex", 0))
return_ECMD_FAILED;
cmd->lockd_gl_disable = 1;
if (!(handle = init_processing_handle(cmd, NULL))) {
log_error("Failed to initialize processing handle.");
return ECMD_FAILED;
}
if (!pvcreate_each_device(cmd, handle, &pp))
ret = ECMD_FAILED;
else {
/* pvcreate_each_device returns with orphans locked */
unlock_vg(cmd, NULL, VG_ORPHANS);
ret = ECMD_PROCESSED;
}
destroy_processing_handle(cmd, handle);
return ret;
}