Skip to content

Wöchentlicher PostgreSQL Newsletter - 13. Dezember 2009

Der Originalartikel befindet sich unter:

== Wöchentlicher PostgreSQL Newsletter - 13. Dezember 2009 ==

Minor Releases für 8.4.2, 8.3.9, 8.2.15, 8.1.19, 8.0.23 und 7.4.27
werden bald erscheinen. Fertigmachen zum Upgrade!

Robert Haas, ITAGAKI Takahiro, Simon Riggs und Greg Stark haben jetzt
Commit-Rechte im CVS. Glückwunsch!

== PostgreSQL Produkt Neuigkeiten ==

pgpool-II 2.3, ein Connection Pooler und mehr, und
pgpoolAdmin 2.3, ein GUI Administration Werkzeug dafür,
sind erschienen.

== PostgreSQL 8.5 Feature der Woche ==

Exclusion Constraints (von Jeff Davis) erlauben es dir, "unique" Daten
anzugeben, welche ein Feld angeben wie z. B. geometrische Daten, eine
Zeitperiode oder ein Array. Teste 8.5 jetzt.

== PostgreSQL Tipp der Woche ==

pgPool2 ist besser für Loadbalancing in Kombination mit einem
Replikationssystem wie Slony oder Bucardo, und pgBouncer ist besser
für skalierende Connectionpools.

== PostgreSQL Jobs for December ==

== PostgreSQL Lokal ==

Die PUG des New York Capital Districts trifft sich zu ihrem erstem
Meeting am 7. Januar 2010.

Stefan Keller gibt einen Kurs "Einführung in PostGIS/PostgreSQL" an
der HSR Hochschule für Technik Rapperswil (HSR) am 14. Januar 2010
in Zürich, Schweiz.

PGDay Kuba 2010 findet vom 4.-6. Februar 2010 in La Havana an der
Universität für Wissenschaftliche Informatik statt. Teilnehmer aus
Kuba, der Dominikanischen Republik und Venezuela sind eingeladen.

FOSDEM 2010 findet am 6. und 7. Februar 2010 in Brüssel,
Belgien statt.

Die Chemnitzer Linuxtage finden am 13. und 14. März in
Chemnitz, Deutschland statt.

PgCon 2010 findet vom 20.-21. Mai 2010 in Ottawa statt mit Tutorials
vorher am 18. und 19. Der Call for Papers ist offen!

== PostgreSQL in den News ==

Planet PostgreSQL:

Dieser wöchentliche PostgreSQL Newsletter wurde erstellt von David
Fetter und Josh Berkus.

Sende Neuigkeiten und Ankündigungen bis Sonntag, 15 Uhr Pazifischer
Zeit. Bitte sende englische Beiträge an, deutsche an, italienische an

== Reviews ==

Greg Smith reviewed the ACL patch for large objects.

Euler Taveira de Oliveira reviewed the EXPLAIN BUFFERS patch.

Jaime Casanova reviewed Zoltan Boszormenyi's ECPG patch to add SQLDA

Jaime Casanova reviewed Jeff Janes's XLogInsert patch and said he'd
mark it, "ready for committer."  Greg Smith's review was less

ITAGAKI Takahiro reviewed Michael Paquier's patch to allow launching
shell commands from pgbench.

Greg Smith reviewed the documentation in Zdenek Kotala's patch to add
"init" as a pg_ctl option.

== Angewandte Patches ==

Tom Lane committed:

- Add exclusion constraints, which generalize the concept of
  uniqueness to support any indexable commutative operator, not just
  equality.  Two rows violate the exclusion constraint if "row1.col OP
  row2.col" is TRUE for each of the columns in the constraint.  Jeff
  Davis, reviewed by Robert Haas.

- Update time zone data files to tzdata release 2009s: DST law changes
  in Antarctica, Argentina, Bangladesh, Fiji, Novokuznetsk, Pakistan,
  Palestine, Samoa, Syria.  Also historical corrections for Hong Kong.

- Update release notes for releases 8.4.2, 8.3.9, 8.2.15, 8.1.19,
  8.0.23, 7.4.27.

- In pgsql/src/backend/executor/execMain.c, ensure that the result
  tuple of an EvalPlanQual cycle gets materialized before we zap the
  input tuple.  Otherwise, pass-by-reference columns of the result
  slot are likely to contain just references to the input tuple,
  leading to big trouble if the pfree'd space is reused.  Per trouble
  report from Jaime Casanova.  This is a new bug in the recent rewrite
  of EvalPlanQual, so nothing to back-patch.

- In pgsql/contrib/pgbench/pgbench.c, arrange to generate different
  random sequences in the different child processes of a pgbench run,
  when we are using -j > 1 and are emulating threads via fork().
  Otherwise the children all inherit the same random sequence state
  and produce the same random-number sequence.  In the threaded case
  the different threads will share one RNG state, so they will produce
  different subsets of one sequence, which is maybe more correlated
  than a purist would like but will not be "the same".  So we leave
  that case alone.  First noticed by Takahiro Itagaki, and is also
  part of the explanation for the pgbench misbehavior recently
  reported by Jaime Casanova.

- In pgsql/src/backend/utils/adt/varbit.c, fix integer-to-bit-string
  conversions to handle the first fractional byte correctly when the
  output bit width is wider than the given integer by something other
  than a multiple of 8 bits.  This has been wrong since I first wrote
  that code for 8.0 :-(.  Kudos to Roman Kononov for being the first
  to notice, though I didn't use his patch.  Per bug #5237.

- Fix a bug introduced when set-returning SQL functions were made
  inline-able: we have to cope with the possibility that the declared
  result rowtype contains dropped columns.  This fails in 8.4, as per
  bug #5240.  While at it, be more paranoid about inserting binary
  coercions when inlining.  The pre-8.4 code did not really need to
  worry about that because it could not inline at all in any case
  where an added coercion could change the behavior of the function's
  statement.  However, when inlining a SRF we allow sorting, grouping,
  and set-ops such as UNION.  In these cases, modifying one of the
  targetlist entries that the sort/group/setop depends on could
  conceivably change the behavior of the function's statement --- so
  don't inline when such a case applies.

Magnus Hagander committed:

- Update CVS documentation to be more current and add documentation
  about git mirror.  Remove information about cvsup and documentation
  that's more about cvs than our use of cvs.  Backpatch to 8.4 so we
  get the git information up on the website as soon as possible.

- In pgsql/doc/src/sgml/installation.sgml, replace broken link to
  custom local gettext package with one to the main GNU site for

- Fix a couple of broken links to third-party sites.

- In pgsql/doc/src/sgml/installation.sgml, update size references in
  installation instructions to be a bit more up-to-date with current

- In pgsql/src/tools/RELEASE_CHANGES, add notes about updating disk
  and shared memory size information in the documentation when doing
  new major release.

- Allow LDAP authentication to operate in search+bind mode, meaning it
  does a search for the user in the directory first, and then binds
  with the DN found for this user.  This allows for LDAP logins in
  scenarios where the DN of the user cannot be determined simply by
  prefix and suffix, such as the case where different users are
  located in different containers.  The old way of authentication can
  be significantly faster, so it's kept as an option.  Robert Fleming
  and Magnus Hagander

Peter Eisentraut committed:

- Translation updates.

- Add init[db] option to pg_ctl.  pg_ctl gets a new mode that runs
  initdb.  Adjust the documentation a bit to not assume that initdb is
  the only way to run database cluster initialization.  But don't
  replace initdb as the canonical way.  Zdenek Kotala.

- PL/Python array support.  Support arrays as parameters and return
  values of PL/Python functions.

Robert Haas committed:

- In pgsql/contrib/fuzzystrmatch/fuzzystrmatch.c, fix levenshtein with
  costs.  The previous code multiplied by the cost in only 3 of the 7
  relevant locations.  Marcin Mank, slightly adjusted by Robert Haas.

- Export ExplainBeginOutput() and ExplainEndOutput() for auto_explain.
  Without these functions, anyone outside of explain.c can't actually
  use ExplainPrintPlan, because the ExplainState won't be initialized
  properly.  The user-visible result of this was a crash when using
  auto_explain with the JSON output format.  Report by Euler Taveira
  de Oliveira.  Analysis by Tom Lane.  Patch by Robert Haas.

Marc Fournier committed:

- Tag 8.4.2, 8.3.9, 8.2.15, 8.1.19, 8.0.23, 7.4.27.

Andrew Dunstan committed:

- Add YAML to list of EXPLAIN formats. Greg Sabino Mullane, reviewed
  by Takahiro Itagaki.

Bruce Momjian committed:

- Properly define ENABLE_THREAD_SAFETY in conflgure, per suggestion
  from Peter Eisentraut.

ITAGAKI Takahiro committed:

- Add large object access control.  A new system catalog
  pg_largeobject_metadata manages ownership and access privileges of
  large objects.  KaiGai Kohei, reviewed by Jaime Casanova.

- Additional fixes for large object access control.  Use
  pg_largeobject_metadata.oid instead of pg_largeobject.loid to
  enumerate existing large objects in pg_dump, pg_restore, and contrib

== Abgelehnte Patches (bis jetzt) ==

No one was disappointed this week :-)

== Eingesandte Patches ==

KaiGai Kohei sent in three more revision of the ACL patch for large
objects per feedback from Greg Smith and ITAGAKI Takahiro.  ITAGAKI
Takahiro sent in two patches atop them to fix some issues with
pg_restore and contrib.

ITAGAKI Takahiro sent in two more revisions of the VACUUM FULL patch.

ITAGAKI Takahiro sent in a patch to fix a bug in the auto_explain

Michael Paquier sent in two more revisions of the patch to add shell
commands to pgbench, per review from Greg Smith.  ITAGAKI Takahiro
sent in another one.

Magnus Hagander sent in a doc patch for CVS and git.

Zdenek Kotala sent in two more revisions of the pg_ctl init extension
per review from Greg Smith.

ITAGAKI Takahiro sent in three more revisions of the EXPLAIN BUFFERS
patch per feedback from Euler Taveira de Oliveira.  Robert Haas sent
in some doc and idiom changes atop those, then ITAGAKI Takahiro sent
in yet another revision.

Fujii Masao sent in a patch for the January 2010 commitfest which
adds XLOG UNLOGGED to how WALs are handled.

Marcin Mank sent in and Robert Haas reviewed and fixed a patch to fix
a bug in Levenshtein distance calculation in contrib/fuzzystrmatch.

Bruce Momjian sent in a patch to install PL/pgsql by default.  Tom
Lane reviewed same.

Joachim Wieland sent in another revision of the patch to add payloads

Zdenek Kotala sent in a WIP patch to add dtrace probes for the memory
manager.  Bernd Helmle sent in an updated revision which splits it
into executor and slru parts.

Robert Haas sent in a patch to fix an issue where ALTER TABLE ...
RENAME COLUMN can change a VIEW.

Andreas Freund sent in another revision of the patch to make tsearch's
parser more efficient.

Peter Eisentraut sent in a WIP patch for PL/Python3.

Robert Haas sent in a fix atop the EXPLAIN YAML patch.

Simon Riggs sent in another revision of the Hot Standby patch.

KaiGai Kohei sent in two reworked patched for an ACE (Access Control
Extension) framework.

James Pye sent in another revision of the PL/Python3 patch.


No Trackbacks


Display comments as Linear | Threaded

No comments

Add Comment

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
E-Mail addresses will not be displayed and will only be used for E-Mail notifications.
To leave a comment you must approve it via e-mail, which will be sent to your address after submission.
Form options