PgSql

crash-recovery replay of CREATE TABLESPACE is broken in HEAD
Heikki Linnakangas <heikki.linnakangas(a)enterprisedb.com> writes: Maybe you should check that it points to the right location? Or drop and recreate the symlink, and ignore failure at mkdir. More specifically, ignore EEXIST failure when replaying mkdir. Anything else is still a problem. regards, tom lan... 18 Jul 2010 13:47

Fix for libpq compile
Bruce Momjian <bruce(a)momjian.us> writes: I have applied the attached check, but without SO_PEERCRED I can't report the username which failed. Please adjust this if necessary. AFAICT that should have been using the uid variable, not any of the platform-specific things. Apparently this was not tested before c... 18 Jul 2010 13:47

crash-recovery replay of CREATE TABLESPACE is brokenin HEAD
On 18/07/10 08:22, Bruce Momjian wrote: The bug is that we can't replay mkdir()/symlink() and assume those will always succeed. I looked at the createdb redo code and it basically drops the directory before creating it. The tablespace directory/symlink setup is more complex, so I just wrote the attac... 18 Jul 2010 02:00

crash-recovery replay of CREATE TABLESPACE isbroken in HEAD
Tom Lane wrote: I managed to crash the executor in the tablespace.sql test while working on a 9.1 patch, and discovered that the postmaster fails to recover from that. The end of postmaster.log looks like LOG: all server processes terminated; reinitializing LOG: database system was interrupted; la... 20 Jul 2010 14:40

Review: Row-level Locks & SERIALIZABLE transactions,postgres vs. Oracle
On 7/17/10 12:09 PM, Kevin Grittner wrote: Joe Conway wrote: Should I be installing Florian's patch in addition to yours when I start testing? There's some manual fix-up needed, primarily because we need to differentiate between SERIALIZABLE and REPEATABLE READ isolation levels, and there... 17 Jul 2010 16:19

Review: Row-level Locks & SERIALIZABLE transactions, postgres vs. Oracle
Joe Conway wrote: Should I be installing Florian's patch in addition to yours when I start testing? There's some manual fix-up needed, primarily because we need to differentiate between SERIALIZABLE and REPEATABLE READ isolation levels, and therefore replaced the IsXactIsoLevelSerializable macro with I... 17 Jul 2010 15:13

Broken due to CVS branching? .bki has wrong info for build
On Jul 17, 2010, at 12:44 PM, Selena Deckelmann <selenamarie(a)gmail.com> wrote: Hi! I pulled the latest from the git repo, I got this error on initdb: creating template1 database in testdb/base/1 ... initdb: input file "/usr/local/pg90/share/postgresql/postgres.bki" does not belong to PostgreSQL 9... 17 Jul 2010 15:13

[HACKERS] Broken due to CVS branching? .bki has wrong info for build
Hi! I pulled the latest from the git repo, I got this error on initdb: creating template1 database in testdb/base/1 ... initdb: input file "/usr/local/pg90/share/postgresql/postgres.bki" does not belong to PostgreSQL 9.0devel The problem was having '9.1' instead of '9.0' in the first line of the .bki. Any... 17 Jul 2010 14:08

Review: Row-level Locks & SERIALIZABLE transactions,postgres vs. Oracle
On 07/17/2010 09:25 AM, Kevin Grittner wrote: I was concerned about its interaction with the other serializable patch (by myself and Dan Ports), so I also combined the patches and tested. Florian's pgbench test did expose bugs in the *other* patch, which I then fixed in the combined setting. There was sti... 17 Jul 2010 14:08

Review: Row-level Locks & SERIALIZABLE transactions, postgres vs. Oracle
================= Submission review ================= * Is the patch in context diff format? Yes. * Does it apply cleanly to the current CVS HEAD? Yes. * Does it include reasonable tests, necessary doc patches, etc? There is one pgbench test which shows incorrect behavior without the patch a... 17 Jul 2010 13:03