1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-24 21:34:56 +03:00
Commit Graph

96 Commits

Author SHA1 Message Date
Mathieu Parent
bb3d6698e9 Move platform-specific code to common/system_*
This removes #ifdef AIX and ease the addition of new platforms.

(This used to be ctdb commit 2fd1067a075fe0e4b2a36d4ea18af139d03f17bf)
2011-12-06 11:57:11 +11:00
Mathieu Parent
4ce585cd0e Remove zero-length gnu_printf format string in ctdb_daemon.c (gcc warning)
server/ctdbd.c: In function ‘main’:
server/ctdb_daemon.c:943:7: warning: zero-length gnu_printf format string [-Wformat-zero-length]

(This used to be ctdb commit e6d1dd3ec4a078e5f32bc52a4a9e4b7d9a2e2d16)
2011-12-06 11:56:18 +11:00
Volker Lendecke
5a1da0ac55 Add CTDB_CONTROL_CHECK_SRVID
(This used to be ctdb commit ad64ef2c40a2a12b37dbf39142e95c6781c2fc3b)
2011-11-30 09:02:26 +11:00
Ronnie Sahlberg
0e79b2d1e8 Record Fetch Collapse: Collapse multiple fetch request into one single request.
When multiple clients fetch the same record concurrently, send only one single
fetch across the network and deferr all other fetches locally.
This improves performance for hot records and reduces cpu load on ctdb.

(This used to be ctdb commit 82d6946ad8b3348e8b9d3d971f24925ade02d1be)
2011-11-08 16:08:28 +11:00
Ronnie Sahlberg
9729d3e339 ReadOnly: Check the readonly flag instead of whether the tdb pointer is NULL or not
(This used to be ctdb commit 01314c2cb3a480917d6a632b83c39f0a48bba0e7)
2011-08-23 10:41:52 +10:00
Ronnie Sahlberg
de7c3de0a2 ReadOnly: clear out the tracking record once a revoke is completed
(This used to be ctdb commit 7af255551f058d1f6bfdd38ca603e7a19d1bb7ba)
2011-08-23 10:35:56 +10:00
Ronnie Sahlberg
38e8964910 ReadOnly: Add handlign of readonly requests readwrite requests, delegations and revoking of delegation to the processing loop for CALL requests coming in from a local client via domain socket
(This used to be ctdb commit e7cbf5b5d03cc26a73a92066a651f8eab73624b8)
2011-08-23 10:32:42 +10:00
Michael Adam
4cca8876e2 daemon: fill ctdb->ctdbd_pid early
(This used to be ctdb commit 3da1e2e30bf34622f08e6ecd5b8fe55684e5007a)
2011-03-14 13:35:50 +01:00
Michael Adam
ee44c23cd5 daemon: correctly end a running trans3_commit if the client disconnects.
(This used to be ctdb commit 9e0898db6df52d9bc799dd87bfea8c72d5f70ba0)
2011-02-24 10:35:25 +01:00
Ronnie Sahlberg
d903473d82 We can not always rely on the recovery daemon pinging us in a timely manner
so we need a "ticker" in the main ctdbd daemon too to ensure we get at least one event to process every second.

This will improve the accuracy of "Time jumped" messages and remove false positives when the recovery daemon is "slow".

(This used to be ctdb commit 70154e5e19e219de086b2995d41e8f6e069ee20d)
2011-01-14 09:47:44 +11:00
Ronnie Sahlberg
ea0df6d882 Revert scheduling back to use real-time processes
Revert this patch:
commit 482c302d46e2162d0cf552f8456bc49573ae729d

We may need to use real-time processes for the main daemon and the recovery daemon to handle the cases where systems come under very high loads.

(This used to be ctdb commit 08bef9dcab6e4da15fc783f8624e5ed09aa060b5)
2011-01-11 07:40:35 +11:00
Ronnie Sahlberg
83e68b62dd delay loading the public ip address file until after we have started the transport and discovered ouw own pnn number
(This used to be ctdb commit 1b57fc866fc836b5dbd3ef7b646e5a0f4280e81e)
2010-11-10 14:55:24 +11:00
Ronnie Sahlberg
5f76f3c0e2 Add a new tunable : DisableIPFailover that when set to non 0
will stopp any ip reallocations at all from happening.

(This used to be ctdb commit d8d37493478a26c5f1809a5f3df89ffd6e149281)
2010-11-10 14:55:24 +11:00
Ronnie Sahlberg
db8cb31d8b during shutdown there is a window after we have stopped TCP and disconnected from all other nodes but before we have stopped all processing.
During this window we may still hit asynchronous events that will fail because we can not send/receive packets from other nodes.

These messages are logged as ... Transport is DOWN. To help indicate that they are benign messages related to the process of shutting down.

These messages spam the syslog during normal shutdown, so this patch will drop the loglevel of these messages to DEBUG, so that they will not appear in or spam the syslog.

(This used to be ctdb commit 8275d265d2ae19b765e30ecf18f6b6319b6e6453)
2010-10-28 13:41:08 +11:00
Ronnie Sahlberg
5ef29f9f25 Update latency countes to show min/max and average
(This used to be ctdb commit 1919e949af4641ffe919123e44b02fb87c13ab9f)
2010-10-11 15:12:24 +11:00
Ronnie Sahlberg
9f66a93f12 Add rolling statistics that are collected across 10 second intervals.
Add a new command "ctdb stats [num]" that prints the [num] most recent statistics intervals collected.

(This used to be ctdb commit e6e16fcd5a45ebd3739a8160c8fb5f44494edb9e)
2010-09-29 12:14:45 +10:00
Ronnie Sahlberg
39c367a68f Create macros to update the statistics counters and use these macros
everywhere instead of manipulating the coutenrs directly.

(This used to be ctdb commit 2e648df890e5713bc575965d87937827b068d0d7)
2010-09-29 12:14:24 +10:00
Ronnie Sahlberg
c6e20a06c7 set up a handler to catch and log debug messages from the tevent layer
(This used to be ctdb commit fdb4c02f595fa207310a9a48da3fefd653fa9e4b)
2010-09-28 08:30:26 +10:00
Ronnie Sahlberg
ac335e3e5d run the "init" event before we freeze the databases
so that we can read from databases during this event

(This used to be ctdb commit 6c93bf5a1219617bfb39b093aee3200c74c2c61a)
2010-08-25 08:35:24 +10:00
Ronnie Sahlberg
e8ffb0d8a4 We use eventloop nesting in a couple of places, notably the sync
parts of the recovery daemon.

Initialize all event contexts to allow nesting

(This used to be ctdb commit 5bf6bd5e7f33aabbeb7b9707716ef99cf471e590)
2010-08-18 10:11:59 +10:00
Rusty Russell
f93440c4b7 event: Update events to latest Samba version 0.9.8
In Samba this is now called "tevent", and while we use the backwards
compatibility wrappers they don't offer EVENT_FD_AUTOCLOSE: that is now
a separate tevent_fd_set_auto_close() function.

This is based on Samba version 7f29f817fa.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>


(This used to be ctdb commit 85e5e760cc91eb3157d3a88996ce474491646726)
2010-08-18 09:16:31 +09:30
Rusty Russell
7061ceffd8 Report client for queue errors.
We've been seeing "Invalid packet of length 0" errors, but we don't know
what is sending them.  Add a name for each queue, and print nread.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>


(This used to be ctdb commit e6cf0e8f14f4263fbd8b995418909199924827e9)
2010-07-01 23:08:49 +10:00
Ronnie Sahlberg
fa618aa66a add additional logging when tdb_chainunlock() fails
so we can see where it was called from when it fails

(This used to be ctdb commit 0c091b3db6bdefd371787d87bc749593ea8e3c76)
2010-06-09 14:37:16 +10:00
Rusty Russell
d5f6026a22 libctdb: reorganize headers: remove ctdb.h, add ctdb_client.h and ctdb_protocol.h
ctdb_client.h is the existing internal client interface (which was mainly
in ctdb.h), and ctdb_protocol.h is the information needed for the wire
protocol only.

ctdb.h will be the new, shiny, libctdb API.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

(This used to be ctdb commit 4bba6b8cd47b352f98d41f9f06258d5ac3c9adef)
2010-05-20 15:18:30 +09:30
Ronnie Sahlberg
eeeb89e3e2 Reduce the loglevel for two log messages for Registering and Deregistering server ids.
BZ61890

(This used to be ctdb commit 944434eb6420774e42e58984c6ddaa326a6853bd)
2010-03-30 11:57:25 +11:00
Stefan Metzmacher
3419e9c4dd server: add "setup" event
This is needed because the "init" event can't use 'ctdb' commands.

metze

(This used to be ctdb commit 1493436b6b24eb05a23b7a339071ad85f70de8f4)
2010-02-23 10:38:49 +01:00
Andrew Tridgell
f23b82b58c ctdb: when we fill the client packet queue we need to drop the client
We can't just drop packets to the list, as those packets could be part
of the core protocol the client is using. This happens (for example)
when Samba is doing a traverse. If we drop a traverse packet then
Samba hangs indefinately. We are better off dropping the ctdb socket
to Samba.

(This used to be ctdb commit a7a86dafa4d88a6bbc6a71b77ed79a178fd802a6)
2010-02-04 15:37:59 +11:00
Stefan Metzmacher
fd06167caa server: add "init" event
This is needed because the "startup" event runs after the initial recovery,
but we need to do some actions before the initial recovery.

metze

(This used to be ctdb commit e953808449c102258abb6cba6f4abf486dda3b82)
2010-01-20 09:44:36 +01:00
Ronnie Sahlberg
4c722fe34c fix a conflict in the merge from rusty
Merge commit 'rusty/ctdb-no-setsched'

Conflicts:

	server/ctdb_vacuum.c

(This used to be ctdb commit b4365045797f520a7914afdb69ebd1a8dacfa0d9)
2009-12-17 08:18:04 +11:00
Rusty Russell
af2613e16f ctdb: use mlockall, cautiously
We don't want ctdb stalling due to paging; this can be far worse than
scheduling delays.  But if we simply do mlockall(MCL_FUTURE), it
increases the risk that mmap (ie. tdb open) or malloc will fail,
causing us to abort.

This patch is a compromise: we mlock all current pages (including
10k of future stack for expansion) and then relock when a client
asks us to open a TDB.  We warn, but don't exit, if it fails.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>


(This used to be ctdb commit 82f778e85440bc713d3f87c08ddc955d3cfce926)
2009-12-16 20:57:20 +10:30
Rusty Russell
c488ba440a Remove RT priority, use niceness.
1) It's buggy.  Code needs to be carefully written (ie. no busy
   loops) to handle running with it, and we fork and run scripts.[1]

2) It makes debugging harder.  If ctdbd loops (as has happened recently)
   it can be extremely hard to get in and see what's happening.  We've already
   seen the valgrind hacks.

3) We have seen recent scheduler problems.  Perhaps they are unrelated,
   but removing this very unusual setup is unlikely to hurt.

4) It doesn't make anything faster.  Under all but the most perverse of
   circumstances, 99% of the cpu gives the same performance as 100%, and
   we will always preempt normal processes anyway.

[1] I made this worse in 0fafdcb8d353 "eventscript: fork() a child for
    each script" by removing the switch_from_server_to_client() which
    restored it, but even that was only for monitor scripts.  Others were
    run with RT priority.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>


(This used to be ctdb commit 482c302d46e2162d0cf552f8456bc49573ae729d)
2009-12-16 19:26:22 +10:30
Stefan Metzmacher
94bc40307a server: Use tdb_check to verify persistent tdbs on startup
Depending on --max-persistent-check-errors we allow ctdb
to start with unhealthy persistent databases.

The default is 0 which means to reject a startup with
unhealthy dbs.

The health of the persistent databases is checked after each
recovery. Node monitoring and the "startup" is deferred
until all persistent databases are healthy.

Databases can become healthy automaticly by a completely
HEALTHY node joining the cluster. Or by an administrator
with "ctdb backupdb/restoredb" or "ctdb wipedb".

metze

(This used to be ctdb commit 15f133d5150ed1badb4fef7d644f10cd08a25cb5)
2009-12-16 08:06:10 +01:00
Stefan Metzmacher
9a96ae0c97 server: only do the mkdir() calls for db_directory* once at the start
metze

(This used to be ctdb commit f30f33685db50860b6cd6fd1b6bdc3066620a78f)
2009-12-16 08:03:56 +01:00
Volker Lendecke
a0d9bd3c13 Run only one event for each epoll_wait/select call
This might be a bit less efficient, but experience in winbind has shown that
event callbacks can trigger changes in the socket state in very hard to
diagnose ways.

(This used to be ctdb commit a78b8ea7168e5fdb2d62379ad3112008b2748576)
2009-12-10 07:52:16 +11:00
Ronnie Sahlberg
fab11acc65 lower the loglevel for the message that a client has attached through a domian socket
(This used to be ctdb commit de9e5236b20d70eac5ed29991703d6d25a103963)
2009-12-02 14:51:57 +11:00
Ronnie Sahlberg
6bad4a4836 Add a proper function to process a process-exist control in the daemon.
This controls is only used by samba when samba wants to check if a subrecord held by a <node-id>:<smbd-pid> is still valid or if it can be reclaimed.

If the node is banned or stopped, we kill the smbd process and return that the process does not exist to the caller. This allows us to recover subrecords from stopped/banned nodes where smbd is hung waiting for the databases to thaw.

bz58185

(This used to be ctdb commit 157807af72ed4f7314afbc9c19756f9787b92c15)
2009-12-02 13:58:27 +11:00
Ronnie Sahlberg
1c7de7a2ed Add a double linked list to the ctdb_context to store a mapping between client pids and client structures.
Add the mapping to the list everytime we accept() a new client connection
and set it up to remove in the destructor when the client structure is freed.

(This used to be ctdb commit f75d379377f5d4abbff2576ddc5d58d91dc53bf4)
2009-12-02 13:41:04 +11:00
Ronnie Sahlberg
bf27dc2d53 Use the PID we pick up from the domain socket when a client connects
and store this in the client structure.

There is no need to rely on the hack that samba sends some special message
handle registrations that encodes the pid in the srvid any more.

This might not work on AIX since I recall some issues to get the pid in
this way on that platform.

(This used to be ctdb commit b4a7efa7e53e060a91dea0e8e57b116e2aeacebf)
2009-12-02 13:17:12 +11:00
Ronnie Sahlberg
e33722a569 start the syslog child a little later, after we have forked and detached from the local shell
(This used to be ctdb commit 9ffd54b73c0d64b67e8e736d7cb54490e77ffa78)
2009-10-30 19:39:11 +11:00
Ronnie Sahlberg
4d40b86805 for debugging
add a global variable holding the pid of the main daemon.
change the tracking of time() in the event loop to only check/warn when called from the main daemon

(This used to be ctdb commit a10fc51f4c30e85ada6d4b7347b0f9a8ebc76637)
2009-10-27 13:18:52 +11:00
Ronnie Sahlberg
86d1b4c465 Add a mechanism where we can register notifications to be sent out to a SRVID when the client disconnects.
The way to use this is from a client to :
1, first create a message handle and bind it to a SRVID
   A special prefix for the srvid space has been set aside for samba :
   Only samba is allowed to use srvid's with the top 32 bits set like this.
   The lower 32 bits are for samba to use internally.

2, register a "notification" using the new control :
                    CTDB_CONTROL_REGISTER_NOTIFY         = 114,
   This control takes as indata a structure like this :
struct ctdb_client_notify_register {
        uint64_t srvid;
        uint32_t len;
        uint8_t notify_data[1];
};

srvid is the srvid used in the space set aside above.
len and notify_data is an arbitrary blob.
When notifications are later sent out to all clients, this is the payload of that notification message.

If a client has registered with control 114 and then disconnects from ctdbd, ctdbd will broadcast a message to that srvid to all nodes/listeners in the cluster.

A client can resister itself with as many different srvid's it want, but this is handled through a linked list from the client structure so it mainly designed for "few notifications per client".

3, a client that no longer wants to have a notification set up can deregister using control
                    CTDB_CONTROL_DEREGISTER_NOTIFY       = 115,
which takes this as arguments :
struct ctdb_client_notify_deregister {
        uint64_t srvid;
};

When a client deregisters, there will no longer be sent a message to all other clients when this client disconnects from ctdbd.

(This used to be ctdb commit f1b6ee4a55cdca60f93d992f0431d91bf301af2c)
2009-10-23 15:24:51 +11:00
Ronnie Sahlberg
a92ba7f729 lower the debug levels for the "create FD messages" so we dont fill up the logs.
(This used to be ctdb commit 87146db2769c2ec494813685bf9cec0d2a6336c3)
2009-10-21 15:26:24 +11:00
Ronnie Sahlberg
9b8c72c446 When clients have blocked, perhaps because the node is banned or stopped and the client is blocked trying to tdb_fetch() a record, make sure we dont queue up too many REQ_MESSAGES.
Add a new tunable to control the maximum queue size we allow to a blocked client before we start discarding REQ_MESSAGES instead of queueing them for delivery.

    This avoids having queued up very very large number of MESSAGES that samba semds
     between eachother to nodes that are blocked/banned/stopped for extended periods
    .

(This used to be ctdb commit f76d6fed8f9630450263b9fa4b5fdf3493fb1e11)
2009-10-21 15:20:55 +11:00
Ronnie Sahlberg
9de3652380 add logging everytime we create a filedescriptor in the main ctdb daemon
so we can spot if there are leaks.

plug two leaks for filedescriptors related to when sending ARP fail
and one leak when we can not parse the local address during tcp connection establish

(This used to be ctdb commit ddd089810a14efe4be6e1ff3eccaa604e4913c9e)
2009-10-15 11:24:54 +11:00
Ronnie Sahlberg
73c0adb029 initial attempt at freezing databases in priority order
(This used to be ctdb commit e8d692590da1070c87a4144031e3306d190ebed2)
2009-10-12 12:08:39 +11:00
Michael Adam
4cd06a330e Fix persistent transaction commit race condition.
In ctdb_client.c:ctdb_transaction_commit(), after a failed
TRANS2_COMMIT control call (for instance due to the 1-second
being exceeded waiting for a busy node's reply), there is a
1-second gap between the transaction_cancel() and
replay_transaction() calls in which there is no lock on the
persistent db. And due to the lack of global state
indicating that a transaction is in progress in ctdbd, other nodes
may succeed to start transactions on the db in this gap and
even worse work on top of the possibly already pushed changes.
So the data diverges on the several nodes.

This change fixes this by introducing global state for a transaction
commit being active in the ctdb_db_context struct and in a db_id field
in the client so that a client keeps track of _which_ tdb it as
transaction commit running on. These data are set by ctdb upon
entering the trans2_commit control and they are cleared in the
trans2_error or trans2_finished controls. This makes it impossible
to start a nother transaction or migrate a record to a different
node while a transaction is active on a persistent tdb, including
the retry loop.

This approach is dead lock free and still allows recovery process
to be started in the retry-gap between cancel and replay.
Also note, that this solution does not require any change in the
client side.

This was debugged and developed together with
Stefan Metzmacher <metze@samba.org> - thanks!

Michael

(This used to be ctdb commit f88103516e5ad723062fb95fcb07a128f1069d69)
2009-07-29 11:12:39 +10:00
Ronnie Sahlberg
d4b30b34aa dont even try to send a message from the main daemon if the transport is down
(This used to be ctdb commit 9a2c4c3ed09ac9ea781d06999d11e5c3b5b4a97a)
2009-06-30 12:09:28 +10:00
Ronnie Sahlberg
4259156050 dont remove the socket when the dameon stops. This can race if the
service is immediately restarted

(This used to be ctdb commit b18356764cd49d934eab901e596bb75c6e3ecdf8)
2009-05-29 18:16:13 +10:00
Sumit Bose
2fcedf6dac add missing checks on so far ignored return values
Most of these were found during a review by Jim Meyering <meyering@redhat.com>

(This used to be ctdb commit 3aee5ee1deb4a19be3bd3a4ce3abbe09de763344)
2009-05-21 11:22:21 +10:00
root
bfea570af4 when tracking the ctdb statistics, only decrement num_clients and pending_calls IFF the counter is >0
Otherwise there is the chance that we will reset the statistics after the counter has been incremented (client connects) to zero   and when the client disconnects we decrement it to a negative number.

this is a pure cosmetic patch with no operational impact to ctdb

(This used to be ctdb commit 72f1c696ee77899f7973878f2568a60d199d4fea)
2009-05-01 12:30:26 +10:00