IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
This allows to differentiate between the two database models.
ctdb_db_persistent() - replicated and permanent
ctdb_db_volatile() - distributed and temporary
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Mon Jun 26 20:10:38 CEST 2017 on sn-devel-144
... and there is no need to find out if mutexes are enabled.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
There is no need for with_jenkinshash and with_mutexes flags, since the
tdb_flags are now calculated based on database type.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
The tdb open flags should be calculated based on the database type and
ctdb tunables.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12856
This stops logger reading from stdin.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Sat Jun 24 14:37:48 CEST 2017 on sn-devel-144
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
If we drop public IPs because CTDB is in recovery for too long, then
avoid spamming logs "Trigger takeoverrun" every second.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
This can be used later in the main_loop to avoid the local ip check.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
This simplifies the code and avoids complicated conditions.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
If the recovery mode is already set to the expected value, there is
nothing to do.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Mon Jun 19 19:56:22 CEST 2017 on sn-devel-144
Currently, every lock helper will log a message if it cannot get a lock.
This can spam the logs and overwhelm syslog if there are hundreds of
lock helpers waiting for contended record.
Instead keep track of the record for which we have already logged once
with specific timeout interval. If we get timeout interval larger than
the previously logged interval, then log again once. This will reduce
the amount of logs for contended records to a single log entry per 10
seconds per record.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Persistent databases are now always recovered by sequence number. So
there is no need to keep the empty records in the database since they
will never be recovered record-by-record using RSN.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Sat Jun 17 16:47:55 CEST 2017 on sn-devel-144
Persistent databases are now always recovered by sequence number. So
there is no need to keep the empty records in the database since they
will never be recovered record-by-record using RSN.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
This avoids spamming the logs during recovery at NOTICE level.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Jun 13 13:22:09 CEST 2017 on sn-devel-144
When thousands of connections are being killed the logs are flooded
with information about connections that should be killed. When some
connections are not killed then the number not killed is printed.
This is the wrong way around! When debugging "fail-back" problems, it
is important to know details of connections that were *not* killed.
It is almost never important to know the full list of all connections
that were *supposed* to be killed.
Instead, print a summary showing how many connections of the total
were killed. If any were not killed then print a list of remaining
connections.
Update unit tests: infrastructure for fake TCP connections, existing,
test cases, add new test cases.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
The non-blocking logging variants can currently silently drop messages
when the socket queue fills.
In this case, count the number of dropped messages and attempt to log
a message about dropped log messages when the next message is logged.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Failures in startup/shutdown/releaseip/takeip are currently
incorrectly ignored.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12837
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Wed Jun 7 20:19:06 CEST 2017 on sn-devel-144
This wraps the entire async computation for setting and removing message
handlers instead of calling multiple sync calls.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
This implements the async version of the traverse code in the ctdb tool
for catdb command.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
This function only traverses the database on local node.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Signed-off-by: Richard Sharpe <realrichardsharpe@gmail.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Jun 7 09:22:29 CEST 2017 on sn-devel-144
Setting CTDB_LOGGING to syslog:nonblocking or syslog:udp will cause
ctdbd to crash at startup due to NULL pointer dereference.
Refactoring in commit c9124a001f
introduced this regression.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12814
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Thu Jun 1 15:26:19 CEST 2017 on sn-devel-144
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue May 30 08:05:56 CEST 2017 on sn-devel-144
This allows to pass data to a child process via stdin.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12802
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Fri May 26 05:24:34 CEST 2017 on sn-devel-144
The "Number of nodes:" header should only be displayed when "all" is
specified. This is how the command behaved in Samba <= 4.4.
Printing the number of nodes is not helpful and is rather confusing in
the default case where only the status of the current node is printed.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12802
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Exit code should only reflect current or specified nodes too.
Drop an unwanted call to get_nodemap() that overwrites the previously
calculated node map.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12802
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
This allows to mix CTDB major versions in a single cluster.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Wed May 24 21:06:28 CEST 2017 on sn-devel-144
Extend CTDB_REQ_KEEPALIVE packet to include version and uptime. If CTDB
versions do not match shutdown ctdb.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
It's not used any more as the version headers are generated from waf.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Generate version headers from waf instead of separate shell script.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12697
During revoking readonly delegations, if one of the nodes disappears,
then there is no point re-trying revoking readonly delegation immedately.
The database needs to be recovered before the revoke operation can
succeed.
However, if the revoke is successful, then all the write requests need
to be processed immediately before the read-only requests. This avoids
starving write requests, in case there are read-only requests coming
from other nodes.
In deferred_call_destructor, the result of revoke is not available and
deferred calls cannot be correctly ordered. To correctly order the
deferred calls, process them in revokechild_destructor where the result
of revoke is known.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12697
This reverts commit ad758cb869.
This is an incomplete fix and introduces a regression.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12792
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Mon May 15 08:08:10 CEST 2017 on sn-devel-144
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12792
... instead of expecting the last line to be the matching log entry.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12770
DEBUGLEVEL is defined as a const array reference at compile time.
debug_init() converts that to an allocated array reference at run time.
Since debug_init() is a static function, initialize DEBUGLEVEL via
setup_logging().
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Mon May 8 00:12:25 CEST 2017 on sn-devel-144
Instead of using hopcount as a metric for hot records, use the number
of migrations per second as a metric.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Wed Apr 5 08:35:45 CEST 2017 on sn-devel-144
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12723
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Wed Mar 29 11:07:18 CEST 2017 on sn-devel-144
TMPDIR sometimes goes missing during autobuild. When that happens the
error messages produced by CTDB tests are not very helpful. This
should make it clear.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Mon Mar 20 08:53:02 CET 2017 on sn-devel-144
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12697
During revoking readonly delegations, if one of the nodes disappears, then
there is no point re-trying revoking readonly delegation. The database
needs to be recovered before the revoke operation can succeed. So retry
only after a grace period.
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Fri Mar 17 14:05:57 CET 2017 on sn-devel-144
When ctdb is built in standalone mode, it turned off the python
requirement for submodules by setting Options.options.disable_python
to True before checking for its own (non-optional) python support.
Ad ctdb does not need python for itself or any of the submodules
it is built against, the safest solution seems to be to allow
the python and python-headers checks to not find python.
Signed-off-by: Ian Stakenvicius <axs@gentoo.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12595
It doesn't look like recursion works for appending scripting commands.
Unroll the recursion instead.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Mon Feb 27 03:09:27 CET 2017 on sn-devel-144
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12595
Without this a top-level "make dist" fails to build the CTDB manpages.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
This avoids confusing log messages like:
ctdbd[21635]: releaseip called for an ip '10.1.1.1' that is not a public address
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Fri Feb 24 11:50:36 CET 2017 on sn-devel-144
Redundant releases will be sent to all connected nodes anyway, so this
is no worse. This will facilitate an improvement to avoid sending
releases to nodes with no known IPs.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
As with other controls, processes the errors by hand instead of using
ctdb_client_control_multi_error(). This will make it easier to add
banning credits for failures.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
A simple optimisation to avoid unnecessary communication.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
At the moment this is done as a post-processing step for any nodes
that have no known IPs. However, this doesn't allow testing of
scenarios where there no known IPs on one or more nodes.
Add relevant tests.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
The current code will fetch IP from all connected and all active
nodes, so this can't happen. However, catch it anyway in case the
calling code changes.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
This simplifies error handling and make failures less likely after
send.
This also means that num_nodes is not required in the state.
Also quietly remove unused ev and client from state.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
"ctdb reloadips" use of ipreallocate() can result in a spurious
takeover runs. This can cause a subsequent "ctdb reloadips" to fail
to disable takeover runs (due to there being one already in progress).
There are various possible improvements but a proper fix probably
requires a protocol change. That would mean receiving an ACK for a
takeover run request to indicate that the request will be processes
and then a broadcast to indicate a completed takeover run.
There are various other partial fixes (e.g. de-duping queued takeover
run requests against those in the in-progess queue) and workarounds
(e.g. always do a double ipreallocate() in the tool, which should
absorb the spurious takeover run).
However, this is unlikely to be a real-world problem. Real use cases
should not involve repeatedly reloading the IP configuration.
Instead, work around the problem of flaky tests by manually adding
"ctdb sync" commands to cause extra no-op takeover runs. These should
not add spurious takeover runs and will create synchronisation points
to help avoid the issue.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>