[SCM] CTDB repository - branch 1.0.112 created - ctdb-1.0.111-37-ga14e49a
Ronnie Sahlberg
sahlberg at samba.org
Sun Feb 14 16:54:48 MST 2010
The branch, 1.0.112 has been created
at a14e49ac60e341ff6daa7fd33e4c62415e6d2c0b (commit)
- Log -----------------------------------------------------------------
commit a14e49ac60e341ff6daa7fd33e4c62415e6d2c0b
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Mon Feb 15 10:49:33 2010 +1100
new version 1.0.112-8
commit 31159657d5de72f53574dad11ab843c1279b9f34
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Fri Feb 12 13:19:57 2010 +1100
try to restart rpc-rquotad if it is not running
bz60317
commit 894c310a9530732253f19c29f8351bb8cc0a63b8
Author: Rusty Russell <rusty at rustcorp.com.au>
Date: Fri Feb 12 17:02:56 2010 +1030
Leave sequence number alone when merely migrating records.
(Based on earlier version from Ronnie which modified tdb; this one
is standalone).
When storing records in a tdb that has "automatic seqnum updates"
also check if the actual data for the record has changed or not.
If it has not changed at all, except for possibly the header,
this is likely just a dmaster migration operation in which case
we want to write the record to the tdb but we do not want the tdb
sequence number to be increased.
This resolves the problem of notify.tdb being thrashed under load:
the heuristic in smbd to only reread this when the sequence number
increases (rarely) breaks down.
Before, running nbench --num-progs=512 across 4 nodes, we saw numbers like:
512 1496 118.33 MB/sec execute 60 sec latency 0.00 msec
And turning on latency tracking, this was typical in the logs:
ctdbd: High latency 9380914.000000s for operation lockwait on database notify.tdb
After this commit:
512 2451 143.85 MB/sec execute 60 sec latency 0.00 msec
And no more latency messages...
Signed-off-by: Rusty Russell <rusty at rustcorp.com.au>
commit 16f265e4f2d3a6154dd77215e853a45379aeff80
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 16:49:48 2010 +1100
new version 1.0.112-7
commit 443c4518a0036de257fbe08fb57fc65ffae49bc6
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 12:00:43 2010 +1100
Reduce loglevel for two eventscript related debug messages
commit b0cf567a88deee913ff03f721898af54c2877261
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 11:54:46 2010 +1100
Reducing the log level for a debug message
DEBUG(DEBUG_DEBUG,("pnn %u starting migration of %08x t\
commit 2a9063910e6281f801c3e84f9aacc0589adad812
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 11:49:48 2010 +1100
Reduce the log level for two debug messages
DEBUG(DEBUG_DEBUG,("pnn %u dmaster response %08x\n", ctdb->pnn, ctdb_has
DEBUG(DEBUG_DEBUG,("pnn %u dmaster request on %08x for %u from %u\n",
commit 82172d6a8c6fe519baf064832281d1b24b9983fe
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 11:32:22 2010 +1100
Add a variable CTDB_CHECK_SWAP_IS_NOT_USED="yes"
to control whether or not to check if we are swapping, and produce
useful output into the logfile if we are.
For production systems with dedicated nas-heads we should never swap.
But for developer/test systems we often use smaller nondedicated systems where
we can no longer guarantee that we will not be using swap.
commit ca3050593d5b52997c2d2940f18d53ffee027fde
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 11:19:08 2010 +1100
lower the loglevel for a debug message for redundant releases of public ips
commit 060c4a8eca64b8c90aec23ee67cf02a89da92033
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 11 11:09:39 2010 +1100
Add a new variable : CTDB_NFS_SKIP_KNFSD_ALIVE_CHECK
when set to "yes" this will skip checking if knfsd has hung or not.
bz59626
commit 7530a8885c4df8e0161a2953f2cbfee0ed3421eb
Author: Rusty Russell <rusty at rustcorp.com.au>
Date: Tue Feb 9 12:46:35 2010 +1030
event scripts: add logging for low memory conditions
We should never enter swap; if we do, show the memory state of the machine and the process list. This will help us diagnose what caused the condition before it's too late and the box starts OOM-killing processes.
Signed-off-by: Rusty Russell <rusty at rustcorp.com.au>
commit f382cf76660b48cfc4391247219a442e5aeb1128
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Tue Feb 9 08:33:24 2010 +1100
new version 1.0.112-6
commit d528c05948149c63f4c20c472d59b21a3a1dc529
Author: Andrew Tridgell <tridge at samba.org>
Date: Sun Feb 7 19:02:06 2010 +1100
ctdb: migrate to new dlinklist.h from Samba
commit 64fb20dcfcff5bc067d97bff39e491e40ca45a3f
Author: Andrew Tridgell <tridge at samba.org>
Date: Fri Feb 5 17:11:29 2010 +1100
fixed printing of high latency
commit 82f2ed827caab8999cb3f958c70821a23490fdde
Author: Andrew Tridgell <tridge at samba.org>
Date: Thu Feb 4 14:36:14 2010 +1100
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.
commit ac885788678255f0c8a091f88ee4d440edf818a9
Author: Andrew Tridgell <tridge at samba.org>
Date: Thu Feb 4 14:14:18 2010 +1100
ctdb: move ctdb_io.c to use TLIST_*() macros
This will make large packet queues much more efficient
commit a781f05f9b80e288ae43ca16f109890942937e62
Author: Andrew Tridgell <tridge at samba.org>
Date: Thu Feb 4 14:13:49 2010 +1100
util: added TLIST_*() macros
The TLIST_*() macros are like the DLIST_*() macros, but take both a
head and tail pointer for the list. This means that adding an element
to the end of the list is efficient (it doesn't need to walk the
list).
We should move all uses of the DLIST_*() macros which use
DLIST_ADD_END() to use the TLIST_*() macros instead.
commit 3ec469339e56f5221960ac5b3a69bf8fb553b8f5
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 4 09:54:06 2010 +1100
We only queued up to 1000 packets per queue before we start dropping
packets, to avoid the queue to grow excessively if smbd has blocked.
This could cause traverse packets to become discarded in case the main
smbd daemon does a traverse of a database while there is a recovery
(sending a erconfigured message to smbd, causing an avalanche of unlock
messages to be sent across the cluster.)
This avalance of messages could cause also the tranversal message to be
discarded causing the main smbd process to hang indefinitely waiting
for the traversal message that will never arrive.
Bump the maximum queue length before starting to discard messages from
1000 to 1000000 and at the same time rework the queueing slightly so we
can append messages cheaply to the queue instead of walking the list
from head to tail every time.
commit 4af34ca1bdb4895c2dd1ec280ef14cffab7fc815
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Feb 4 06:37:41 2010 +1100
Drop the debug level for logging fd creation to DEBUG_DEBUG
commit 04e40deac8d0c7edf907135ae81ac961c23135c3
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Tue Feb 2 08:03:37 2010 +1100
Version 1.0.112-5
commit 72d6ae64ab5ece7645b02054d617b71e231d4741
Author: Volker Lendecke <vl at samba.org>
Date: Fri Jan 29 18:21:09 2010 +0100
tdb: fix an early release of the global lock that can cause data corruption
There was a bug in tdb where the
tdb_brlock(tdb, GLOBAL_LOCK, F_UNLCK, F_SETLKW, 0, 1);
(ending the transaction-"mutex") was done before the
/* remove the recovery marker */
This means that when a transaction is committed there is a window where another
opener of the file sees the transaction marker while the transaction committer
is still fully functional and working on it. This led to transaction being
rolled back by that second opener of the file while transaction_commit() gave
no error to the caller.
This patch moves the F_UNLCK to after the recovery marker was removed, closing
this window.
commit 3e2b1839a9f8419eeeb7f22ea5925f6c42f32a65
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Jan 21 18:51:54 2010 +1100
version 1.0.112-4
commit eb68bc6f27f9b10abbd5bc4e4fd62b7af54c9abb
Author: Martin Schwenke <martin at meltin.net>
Date: Thu Jan 21 13:40:03 2010 +1100
onnode: update algorithm for finding nodes file.
2 changes:
* If a relative nodes file is specified via -f or $CTDB_NODES_FILE but
this file does not exist then try looking for the file in /etc/ctdb
(or $CTDB_BASE if set).
* If a nodes file is specified via -f or $CTDB_NODES_FILE but this
file does not exist (even when checked as per above) then do not
fall back to /etc/ctdb/nodes ((or $CTDB_BASE if set). The old
behaviour was surprising and hid errors.
Signed-off-by: Martin Schwenke <martin at meltin.net>
commit cd8b1eb75f430c589e71b2837dab9d83a12bb43e
Author: Martin Schwenke <martin at meltin.net>
Date: Thu Jan 21 13:16:18 2010 +1100
onnode - respect $CTDB_BASE rather than hard-coding /etc/ctdb.
Signed-off-by: Martin Schwenke <martin at meltin.net>
commit 910debaa673319a193374ecc75048e3b8da7564d
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Wed Jan 20 15:14:47 2010 +1100
Version 1.0.112-3
commit 287054d93ce4136d2f3db783ce4b5d3af3d14726
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Wed Jan 20 10:35:02 2010 +1100
source the nfs sysconfig file from the 61.nfstickles script
commit 232f9f40926573fe0f1fb1782280f693d37c38fe
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Wed Jan 13 11:23:23 2010 +1100
new version 1.0.112-2
-----------------------------------------------------------------------
--
CTDB repository
More information about the samba-cvs
mailing list