Steve Singer [Mon, 20 Aug 2018 02:27:51 +0000 (22:27 -0400)]
add missing files for unit test
Steve Singer [Mon, 20 Aug 2018 01:34:23 +0000 (21:34 -0400)]
doc updates
Steve [Sun, 19 Aug 2018 20:20:37 +0000 (13:20 -0700)]
Changes to get slony to compile with MS VC2017 against
PostgreSQL 10 on Windows 10.
Steve Singer [Wed, 8 Aug 2018 16:22:37 +0000 (12:22 -0400)]
2.2.7 function names in export file
Steve Singer [Wed, 8 Aug 2018 15:31:41 +0000 (11:31 -0400)]
indentation fixes
Steve Singer [Wed, 8 Aug 2018 15:20:53 +0000 (11:20 -0400)]
release notes fix
Steve Singer [Mon, 6 Aug 2018 13:51:10 +0000 (09:51 -0400)]
2.2.7
Steve Singer [Mon, 6 Aug 2018 13:50:37 +0000 (09:50 -0400)]
release notes update
Steve Singer [Wed, 8 Aug 2018 15:18:10 +0000 (11:18 -0400)]
Documentation update for SLONY_SHARE_DIR
Steve Singer [Wed, 8 Aug 2018 15:15:11 +0000 (11:15 -0400)]
Have the slony share directory come from the
environment variable SLONY_SHARE_DIR not PGHOME.
Also make this happen with or without PGPORT
Steve Singer [Thu, 7 Jun 2018 01:21:28 +0000 (21:21 -0400)]
Add documentation to the admin guide for the akamai patches
Steve Singer [Mon, 21 May 2018 14:13:21 +0000 (10:13 -0400)]
1. Add an option to slonik to disable the version check
against the slony schema
2. Add an option to make the remote listener use a
READ COMMITTED instead of a serialized deferrable
transaction
3. If slonik was not compiled with pgport then allow
PG_HOME to override/set the location of the share directory
containing the slony .sql files
Patches from Tom Tignor<
[email protected]>.
Steve Singer [Wed, 1 Aug 2018 19:29:28 +0000 (15:29 -0400)]
release notes
Steve Singer [Wed, 1 Aug 2018 19:22:15 +0000 (15:22 -0400)]
update example in the docs (set add sequence)
Steve Singer [Wed, 1 Aug 2018 19:13:46 +0000 (15:13 -0400)]
Fix bug in set add sequence(set id=1, sequences='test\\.seq*');
where the regex esacping for sequences was not working like
it did for tables in 'set add table'.
Steve Singer [Mon, 11 Jun 2018 22:32:58 +0000 (18:32 -0400)]
release notes
Steve Singer [Mon, 11 Jun 2018 22:29:24 +0000 (18:29 -0400)]
fix awk expression so flex 2.6.x or higher is supported
Steve Singer [Thu, 7 Jun 2018 00:47:00 +0000 (20:47 -0400)]
alter slony1_funcs so they can compile on PG11
Steve Singer [Sun, 27 Aug 2017 15:24:58 +0000 (11:24 -0400)]
release notes
Steve Singer [Sat, 24 Mar 2018 02:34:55 +0000 (22:34 -0400)]
fix version check so it doesn't warn about
unsupported PG versions for PG < 11
Thanks Jehan-Guillaume de Rorthais
Steve Singer [Sun, 27 Aug 2017 15:10:56 +0000 (11:10 -0400)]
Fix the version string for PG10
Steve Singer [Sun, 27 Aug 2017 15:10:44 +0000 (11:10 -0400)]
more unit test fixes
Steve Singer [Mon, 21 Aug 2017 02:59:29 +0000 (22:59 -0400)]
Add a sync to make sure the resubscribes have propogated before we start
failing nodes
Steve Singer [Mon, 21 Aug 2017 02:58:49 +0000 (22:58 -0400)]
do_item_review is part of set 2 which is not used in this test.
Remove it from the comparision
Steve Singer [Mon, 14 Aug 2017 02:56:48 +0000 (22:56 -0400)]
adjust unit test.
An earlier change made this unit test loop which wasn't
intended to be committed.
Steve Singer [Mon, 14 Aug 2017 02:55:40 +0000 (22:55 -0400)]
adjust assert in unit test
Steve Singer [Sun, 13 Aug 2017 02:48:38 +0000 (22:48 -0400)]
Fix issue where DDL issued against a non-origin node
can be duplicated.
The code intended to only use the provider that the
event was received from when querying sl_log_script but
not sets however the old queries in the provider structure
from previous syncs were not blanked and those queries were
being issued against the other providers.
Blank the provider query for providers we won't be using and
ignore those providers.
Marc Cousin [Mon, 5 Dec 2016 13:37:02 +0000 (14:37 +0100)]
Support multi-schema in slonik_build_env
merged from master
Steve Singer [Sun, 30 Jul 2017 20:05:39 +0000 (16:05 -0400)]
fix typo
Steve Singer [Sun, 30 Jul 2017 20:05:23 +0000 (16:05 -0400)]
release notes update
Steve Singer [Sat, 29 Jul 2017 01:56:40 +0000 (21:56 -0400)]
postgesql 10
Steve Singer [Sat, 29 Jul 2017 01:56:13 +0000 (21:56 -0400)]
2.2.6 release numbering
Steve Singer [Mon, 10 Jul 2017 02:00:12 +0000 (22:00 -0400)]
Add support for PG10
PostgreSQL 10 does away with the old convention of $MAJOR.$MINOR.$PATCH
The existing version detection parsing code does not deal with the new
convention. Update the configure check, the slonik check and the slon check
to work with the new convention and continue to work with the old one.
Steve Singer [Tue, 25 Apr 2017 12:16:19 +0000 (08:16 -0400)]
failover unit test improvements
Steve Singer [Tue, 25 Apr 2017 12:14:50 +0000 (08:14 -0400)]
Fix segfault introduced by the previous failover
fixes
Steve Singer [Sat, 15 Apr 2017 00:28:58 +0000 (20:28 -0400)]
Further changes related to 'Explicitly disallow createEvent and data changes in the same transaction'
Add in support where the logApply trigger inserst data into a table that has trigger
,which runs on the replica,that then inserts data into another replicated table.
The logApply trigger might have already been called in this transaction but the
logTrigger is then called in the same transaction. We then need to setup
the plans for inserting data into sl_log_1 or sl_log_2
Steve Singer [Mon, 10 Apr 2017 13:40:41 +0000 (09:40 -0400)]
Add unit test for multi-node failover (cascaded) issues
Steve Singer [Mon, 10 Apr 2017 13:38:15 +0000 (09:38 -0400)]
Fix issues with FAILOVER
In particular it is possible when doing a multi-node failover
with a cascaded node for the result after the failover to
have nodes with a self subscription, a row in sl_subscribe
with the provider and receiver equal.
In fixing that issue we also discovered that slon was not
reloading the set origin's into memory following a FAILOVER
command. This could mean that we were processing SYNC events
from a node which without realizing that node now is the origin
for a set.
Steve Singer [Sun, 26 Mar 2017 16:54:46 +0000 (12:54 -0400)]
Explicitly disallow createEvent and data changes in the same transaction.
The same transaction should not have both a row in sl_event and log data
to sl_log_1 or sl_log_2 with the logTrigger. This is not something slonik
would ever do, and the data changes that the logTrigger logs might not get
replicated as part of the event processing(ie for many non SYNC events).
it is also unclear what order the data should be applied in with respect to
the actions of the non-SYNC event.
Steve Singer [Sat, 10 Dec 2016 15:53:40 +0000 (10:53 -0500)]
add release notes for slonik_build_env now supporting
multiple -schema options.
Thanks Marc Cousin <
[email protected]>
Christopher Browne [Tue, 26 Jul 2016 20:05:10 +0000 (16:05 -0400)]
Bug 341 - suppress log trigger/deny when running in 'local' mode
Steve Singer [Mon, 30 May 2016 22:34:34 +0000 (18:34 -0400)]
release 2.2.5 configure update
Steve Singer [Mon, 30 May 2016 01:50:25 +0000 (21:50 -0400)]
update supported pg version
Steve Singer [Sat, 28 May 2016 21:44:24 +0000 (17:44 -0400)]
version 2.2.5
Steve Singer [Sat, 28 May 2016 21:41:42 +0000 (17:41 -0400)]
add supported compatibility for PG 9.5
Steve Singer [Sat, 28 May 2016 16:32:57 +0000 (12:32 -0400)]
Fix GETCONFIGOPTIONBYNAME checking for PG9.6+
Also add the define to config.h.in so it actually gets set
Steve Singer [Mon, 23 May 2016 20:59:11 +0000 (16:59 -0400)]
Add win32/msvc support for 3 argument GetConfigOptionByName
Christopher Browne [Mon, 13 Jul 2015 19:16:07 +0000 (15:16 -0400)]
PostgreSQL HEAD has added a 3-parameter version of GetConfigOptionByName().
Have autoconf detect this, and pass 3 parms if needful.
Steve Singer [Sat, 21 May 2016 22:36:10 +0000 (18:36 -0400)]
release notes update
Steve Singer [Sat, 21 May 2016 22:33:50 +0000 (18:33 -0400)]
Fix compile errors against PG 9.6+
The file parser/keywords.h moved to common/keywords.h
but is not actually needed
Christopher Browne [Mon, 30 Nov 2015 16:37:12 +0000 (11:37 -0500)]
Bug #363 - missing WAIT ON in sample script in docs
Steve Singer [Mon, 9 Nov 2015 21:10:34 +0000 (16:10 -0500)]
Doc sample command change:
exclude the slony schema from the pg_dump for adding a new node
Steve Singer [Mon, 31 Aug 2015 22:45:52 +0000 (18:45 -0400)]
release notes update
Steve Singer [Mon, 31 Aug 2015 22:44:11 +0000 (18:44 -0400)]
Add win32 makefile/config support for PG 9.5
Patch by
[email protected]
Steve Singer [Mon, 19 Jan 2015 01:28:48 +0000 (20:28 -0500)]
win32 uses strtok_s not strtok_r
Steve Singer [Sun, 18 Jan 2015 18:22:28 +0000 (13:22 -0500)]
Fix the patch for bug352 so it compiles against PG 8.3
Steve Singer [Thu, 15 Jan 2015 22:56:05 +0000 (17:56 -0500)]
remove unsupported for PG 9.4 since it has ben released
Steve Singer [Thu, 15 Jan 2015 22:51:52 +0000 (17:51 -0500)]
Preparing for 2.3.4 release
Steve Singer [Thu, 15 Jan 2015 21:27:49 +0000 (16:27 -0500)]
updating release notes
Steve Singer [Mon, 27 Oct 2014 21:01:29 +0000 (17:01 -0400)]
Bug 354
The RebuildListenEntries for a failed node
that provides no subscriptions deletes all listen
entries from that node. The remaining nodes
might still need to listen for these events
if the failover target has them. Otherwise
the FAILOVER_NODE event will wait for this event
but it will never come.
Leave in the listen paths for failed nodes
that don't provide any sets
Steve Singer [Thu, 18 Dec 2014 20:38:08 +0000 (15:38 -0500)]
Bug 356
Apply truncate on replica's as TRUNCATE ONLY.
If the original truncate was a TRUNCATE ONLY then
we also want to just truncate the parent table.
If the original command was an un-restricted truncate
then the truncate trigger will be invoked for each child
table anyway.
Steve Singer [Thu, 15 Jan 2015 20:48:30 +0000 (15:48 -0500)]
release notes update
Steve Singer [Thu, 15 Jan 2015 20:30:21 +0000 (15:30 -0500)]
add disorder test for cleanup interval(bug 350)
Rose Nancy [Thu, 15 Jan 2015 20:29:25 +0000 (15:29 -0500)]
Bug 350
Make the cleanup_interval argument in the parameter file actually do
what one thinks it does, that is set how many seconds between cleanupEvent
runs.
Christopher Browne [Tue, 16 Dec 2014 20:30:40 +0000 (15:30 -0500)]
Bug #349 - case insensitivity with cluster name when processing DDL
Cluster name wasn't properly quoted when processing sequence update, this logic only runs when doing DDL.
Added config change to regression tests to have StudlyCase names to ensure this is noticed in future
Christopher Browne [Mon, 15 Dec 2014 23:21:30 +0000 (18:21 -0500)]
Bug #352 applied to 2.2 STABLE
Steve Singer [Tue, 9 Dec 2014 02:48:06 +0000 (21:48 -0500)]
Fix comment and log message so it no longer
refers to helper threads(which were removed in 2.2)
Steve Singer [Tue, 9 Dec 2014 02:41:18 +0000 (21:41 -0500)]
Fix comment to reflect the actual default value
Patch by Kazuki Uehara <
[email protected]>
Steve Singer [Tue, 23 Sep 2014 13:55:38 +0000 (09:55 -0400)]
Bug 345
The strtok_r on the node list was adding replacing the comma
with a NULL causing only 1 node to get inserted into sl_event.
This mean any nodes that where cascaded only saw 1 of the nodes
in the drop node list.
Call strtok_r on a copy of the list so it can safely be munged.
Also add in a DropNode cluster test to demonstrate test this issue
Steve Singer [Thu, 21 Aug 2014 12:03:15 +0000 (08:03 -0400)]
Include alloca.h
Fix build problems on Solaris/SPARC in slonik due to an
undefined alloca symbol.
Report and fix from Venkata Balaji N
Steve Singer [Mon, 7 Jul 2014 14:31:41 +0000 (10:31 -0400)]
Follow up to bug 338 DROP the ddlScript function
We need to drop the function (if it exists) because
the return type is changing
Steve Singer [Tue, 1 Jul 2014 18:59:12 +0000 (14:59 -0400)]
Preparing for 2.2.3 release
Jan Wieck [Mon, 30 Jun 2014 14:34:29 +0000 (10:34 -0400)]
Continue setting options even if a non-existing one
was encountered. Fixes bug 343.
Steve Singer [Mon, 30 Jun 2014 13:05:40 +0000 (09:05 -0400)]
release notes update
Steve Singer [Mon, 30 Jun 2014 12:37:54 +0000 (08:37 -0400)]
Remove HAVE_POSIX_SIGNALS from config.h
Compiling slony against PG 9.4 gives compiler warnings because the
definition of HAVE_POSIX_SIGNALS has changed. This is defined in pg_config
so we do not really need to redefine it in the config.h slony provides.
We leave the configure check in place to we just don't include it in
the header
Steve Singer [Tue, 13 May 2014 13:49:37 +0000 (09:49 -0400)]
fixing generate load origin node parameter mismatch
Steve Singer [Mon, 12 May 2014 01:37:45 +0000 (21:37 -0400)]
Bug 342 - Fix various failover issues
This commit fixes some related failover issues
1. The query slonik used to get a node list during failover
was ignoring non-failover targets. This meant that we
would never call preFailover on a node that wasn't a failover
target. This node might then still be trying to use
a failed provider.
2. A node (say node 3) might process a FAILOVER_NODE
event from node 2 (the new-origin) but might not
have yet processed the last event from node 1, the
failed node. We need to make sure that the
subscription paths and listen networks are updated
so we don't only listen for events and data from
failed nodes.
Steve Singer [Mon, 12 May 2014 01:35:53 +0000 (21:35 -0400)]
updates/fixes to the failover unit tests
Specify the proper node to the load generator
Steve Singer [Fri, 2 May 2014 18:27:33 +0000 (14:27 -0400)]
add another multi-node failover configuration to the test
David Christensen [Fri, 25 Apr 2014 18:36:32 +0000 (13:36 -0500)]
Fix typo in preamble for regression tests
Steve Singer [Tue, 22 Apr 2014 19:07:16 +0000 (15:07 -0400)]
Release notes update
Steve Singer [Tue, 8 Apr 2014 01:52:19 +0000 (21:52 -0400)]
Bug 338 - Have ddlScript return a bigint since the action_seq is a bigint
Steve Singer [Thu, 17 Apr 2014 13:54:14 +0000 (09:54 -0400)]
Document the EXECUTE ONLY ON list of nodes syntax
Also add a regression test to test this
Rose Nancy [Tue, 11 Mar 2014 14:37:00 +0000 (10:37 -0400)]
fixing Deadlock with application during minor version slony upgrade
Rose Nancy [Tue, 11 Mar 2014 14:10:36 +0000 (10:10 -0400)]
fixing Deadlock with application during minor version slony upgrade
Steve Singer [Wed, 12 Feb 2014 03:11:30 +0000 (22:11 -0500)]
updating version for 2.2.2
Steve Singer [Fri, 7 Feb 2014 20:22:00 +0000 (15:22 -0500)]
2.2.2 version update
Steve Singer [Fri, 7 Feb 2014 13:46:11 +0000 (08:46 -0500)]
release notes update for 2.2.2
Jan Wieck [Wed, 29 Jan 2014 17:11:57 +0000 (12:11 -0500)]
logApply trigger free()s data from wrong cache entry on evict.
Steve Singer [Wed, 5 Feb 2014 23:33:49 +0000 (18:33 -0500)]
Fixing typo in docs
Thanks Jeff Frost
Rose Nancy [Wed, 29 Jan 2014 17:21:34 +0000 (12:21 -0500)]
Clarify namespace setup in documentation
Christopher Browne [Thu, 2 Jan 2014 21:17:17 +0000 (16:17 -0500)]
Fix bug #328 - SGML tagging error
Reported by Christoph Berg
Steve Singer [Sat, 14 Dec 2013 21:41:21 +0000 (16:41 -0500)]
release note update slonik include path fix
Steve Singer [Thu, 12 Dec 2013 14:26:04 +0000 (09:26 -0500)]
Use server includes for pgport when building slonik
When building with --with-pgport slonik needs to include port.h
port.h is normally found under the server include directory but
some distributions (ie debian libpq-dev) also include it in
/usr/include/postgresql
The fix for bug 315 removed the server include paths from
all client programs including slonik. This went too far,
the actual issue with bug 315 was src/backend failing because
it pulled in the client headers.
If we are building with pgport we will pull in the server
include directories for slonik
Steve Singer [Thu, 15 Aug 2013 00:39:37 +0000 (20:39 -0400)]
add in explicit support for PG 9.3
(cherry picked from commit
dc68fcf8d621e9354caa20de5a606f0596e37a2f)
Conflicts:
src/slonik/slonik.c
Steve Singer [Sun, 3 Nov 2013 20:39:06 +0000 (15:39 -0500)]
fixing error handling in unit test
Steve Singer [Sat, 2 Nov 2013 18:56:47 +0000 (14:56 -0400)]
updating version for 2.2.1
Steve Singer [Sat, 2 Nov 2013 01:41:59 +0000 (21:41 -0400)]
release notes update
Jan Wieck [Wed, 30 Oct 2013 16:12:11 +0000 (12:12 -0400)]
Disable the "known failing" disorder tests.
Jan Wieck [Tue, 29 Oct 2013 22:48:25 +0000 (18:48 -0400)]
Folding slon.stop() and coordinator.join(slon) into one loop.
Stopping the slons and joining their processes in separate
loops apparently can cause the regression test to hang under
some rare race conditions.