postgresql/src/test
Alvaro Herrera 533e9c6b06 Avoid serializability errors when locking a tuple with a committed update
When key-share locking a tuple that has been not-key-updated, and the
update is a committed transaction, in some cases we raised
serializability errors:
    ERROR:  could not serialize access due to concurrent update

Because the key-share doesn't conflict with the update, the error is
unnecessary and inconsistent with the case that the update hasn't
committed yet.  This causes problems for some usage patterns, even if it
can be claimed that it's sufficient to retry the aborted transaction:
given a steady stream of updating transactions and a long locking
transaction, the long transaction can be starved indefinitely despite
multiple retries.

To fix, we recognize that HeapTupleSatisfiesUpdate can return
HeapTupleUpdated when an updating transaction has committed, and that we
need to deal with that case exactly as if it were a non-committed
update: verify whether the two operations conflict, and if not, carry on
normally.  If they do conflict, however, there is a difference: in the
HeapTupleBeingUpdated case we can just sleep until the concurrent
transaction is gone, while in the HeapTupleUpdated case this is not
possible and we must raise an error instead.

Per trouble report from Olivier Dony.

In addition to a couple of test cases that verify the changed behavior,
I added a test case to verify the behavior that remains unchanged,
namely that errors are raised when a update that modifies the key is
used.  That must still generate serializability errors.  One
pre-existing test case changes behavior; per discussion, the new
behavior is actually the desired one.

Discussion: https://www.postgresql.org/message-id/560AA479.4080807@odoo.com
  https://www.postgresql.org/message-id/20151014164844.3019.25750@wrigleys.postgresql.org

Backpatch to 9.3, where the problem appeared.
2016-07-15 14:17:20 -04:00
..
examples Update copyright for 2016 2016-01-02 13:33:40 -05:00
isolation Avoid serializability errors when locking a tuple with a committed update 2016-07-15 14:17:20 -04:00
locale Fix omission of -X (--no-psqlrc) in some psql invocations. 2015-12-28 11:46:43 -05:00
mb Revert "Fix broken multibyte regression tests." 2016-01-26 08:29:15 +09:00
modules Provide and use a makefile target to build all generated headers. 2016-07-01 15:09:02 -04:00
perl Fix TAP tests and MSVC scripts for pathnames with spaces. 2016-07-09 16:47:38 -04:00
recovery Finish pgindent run for 9.6: Perl files. 2016-06-12 04:19:56 -04:00
regress Fix parsing NOT sequence in tsquery 2016-07-15 20:01:41 +03:00
ssl Clean the last few TAP suite tmp_check directories. 2016-02-24 23:41:54 -05:00
thread Update copyright for 2016 2016-01-02 13:33:40 -05:00
Makefile Add a test framework for recovery 2016-02-26 16:13:30 -03:00
README Minor tweaks for new src/test/recovery 2016-02-29 18:16:59 -03:00

PostgreSQL tests
================

This directory contains a variety of test infrastructure as well as some of the
tests in PostgreSQL. Not all tests are here -- in particular, there are more in
individual contrib/ modules and in src/bin.

Not all these tests get run by "make check". Check src/test/Makefile to see
which tests get run automatically.

examples/
  Demonstration programs for libpq that double as regression tests via
  "make check"

isolation/
  Tests for concurrent behavior at the SQL level

locale/
  Sanity checks for locale data, encodings, etc

mb/
  Tests for multibyte encoding (UTF-8) support

modules/
  Extensions used only or mainly for test purposes, generally not suitable
  for installing in production databases

perl/
  Infrastructure for Perl-based TAP tests

recovery/
  Test suite for recovery and replication

regress/
  PostgreSQL's main regression test suite, pg_regress

ssl/
  Tests to exercise and verify SSL certificate handling

thread/
  A thread-safety-testing utility used by configure