Release date: 2006-12-05
This release adds many functionality and performance improvements that were requested by users, including:
Query language enhancements including
INSERT/UPDATE/DELETE RETURNING, multirow
VALUES lists, and optional
target-table alias in
Index creation without blocking concurrent
Many query optimization improvements, including support for reordering outer joins
Improved sorting performance with lower memory usage
More efficient locking with better concurrency
More efficient vacuuming
Easier administration of warm standby servers
for tables and indexes
Monitoring, logging, and performance tuning additions
More control over creating and dropping objects
Table inheritance relationships can be defined for and removed from pre-existing tables
COPY TO can copy the
output of an arbitrary
Array improvements, including nulls in arrays
Aggregate-function improvements, including multiple-input aggregates and SQL:2003 statistical functions
A dump/restore using pg_dump is required for those wishing to migrate data from any previous release.
Observe the following incompatibilities:
on by default (Bruce)
This issues a warning if backslash escapes are used in
Change the row constructor
that list elements
will be expanded to a list of their member fields, rather
than creating a nested row type field as formerly
The new behavior is substantially more useful since it
allows, for example, triggers to check for data changes
IF row(new.*) IS DISTINCT FROM
row(old.*). The old behavior is still available by
Make row comparisons follow SQL standard semantics and allow them to be used in index scans (Tom)
Previously, row = and <> comparisons followed the standard but < <= > >= did not. A row comparison can now be used as an index constraint for a multicolumn index matching the row value.
IS [NOT] NULL tests follow
SQL standard semantics
The former behavior conformed to the standard for
simple cases with
IS NOT NULL would return
true if any row field was non-null, whereas the standard
says it should return true only when all fields are
CONSTRAINT affect only one constraint (Kris
In previous releases,
CONSTRAINT modified all constraints with a
matching name. In this release, the schema search path is
used to modify only the first matching constraint. A
schema specification is also supported. This more nearly
conforms to the SQL standard.
for tables, for security reasons (Tom)
As of this release, only a table's owner can create or
modify rules for the table. For backwards compatibility,
REVOKE RULE is still accepted, but it
Array comparison improvements (Tom)
Now array dimensions are also compared.
Change array concatenation to match documented behavior (Tom)
This changes the previous behavior where concatenation would modify the array lower bound.
Make command-line options of postmaster and postgres identical (Peter)
This allows the postmaster to pass arguments to each
backend without using
Note that some options are now only available as
long-form options, because there were conflicting
Deprecate use of postmaster symbolic link (Peter)
postmaster and postgres commands now act identically, with the behavior determined by command-line options. The postmaster symbolic link is kept for compatibility, but is not really needed.
log_duration to output even if the query is
not output (Tom)
In prior releases,
log_duration only printed if the query
appeared earlier in the log.
Most applications should use
HH24 unless they want a 12-hour
This ensures that the converted value is actually
variable (Joachim Wieland)
This variable has been superseded by a more general facility for configuring timezone abbreviations.
Improve cost estimation for nested-loop index scans (Tom)
This might eliminate the need to set unrealistically
small values of
random_page_cost. If you have been using a
please recheck your test cases.
Change behavior of
-t options. (Greg Sabino
page for details.
PQdsplen() to return a useful value
(Martijn van Oosterhout)
PQgetssl() as returning
void *, rather than
SSL * (Martijn van Oosterhout)
This allows applications to use the function without including the OpenSSL headers.
C-language loadable modules must now include a
PG_MODULE_MAGIC macro call for
version compatibility checking (Martijn van
For security's sake, modules used by a PL/PerlU function are no longer available to PL/Perl functions (Andrew)
This also implies that data can no longer be shared between a PL/Perl function and a PL/PerlU function. Some Perl installations have not been compiled with the correct flags to allow multiple interpreters to exist within a single process. In this situation PL/Perl and PL/PerlU cannot both be used in a single backend. The solution is to get a Perl installation which supports multiple interpreters.
xml_valid() will remain
for backward compatibility, but its behavior will change
to do schema checking in a future release.
now at http://www.samse.fr/GPL/ora2pg
Remove contrib modules that have been migrated to
Remove abandoned contrib modules:
Remove QNX and BEOS ports (Bruce)
These ports no longer had active maintainers.
Below you will find a detailed account of the changes between PostgreSQL 8.2 and the previous major release.
Allow the planner to reorder outer joins in some circumstances (Tom)
In previous releases, outer joins would always be evaluated in the order written in the query. This change allows the query optimizer to consider reordering outer joins, in cases where it can determine that the join order can be changed without altering the meaning of the query. This can make a considerable performance difference for queries involving multiple outer joins or mixed inner and outer joins.
Improve efficiency of
IN (list-of-expressions) clauses
Improve sorting speed and reduce memory usage (Simon, Tom)
Improve subtransaction performance (Alvaro, Itagaki Takahiro, Tom)
This leaves extra free space in each table or index page, allowing improved performance as the database grows. This is particularly valuable to maintain clustering.
Increase default values for
Improve locking performance by breaking the lock manager tables into sections (Tom)
This allows locking to be more fine-grained, reducing contention.
Reduce locking requirements of sequential scans (Qingqing Zhou)
Reduce locking required for database creation and destruction (Tom)
SELECT already honored
Improve planning of constant
WHERE clauses, such as a condition
that depends only on variables inherited from an outer
query level (Tom)
Protocol-level unnamed prepared statements are
re-planned for each set of
BIND values (Tom)
This improves performance because the exact parameter values can be used in the plan.
Speed up vacuuming of B-Tree indexes (Heikki Linnakangas, Tom)
Avoid extra scan of tables without indexes during
Improve multicolumn GiST indexing (Oleg, Teodor)
Remove dead index entries before B-Tree page split (Junji Teramoto)
Allow a forced switch to a new transaction log file (Simon, Tom)
This is valuable for keeping warm standby slave
servers in sync with the master. Transaction log file
switching now also happens automatically during
pg_stop_backup(). This ensures that all
transaction log files needed for recovery can be
Add WAL informational functions (Simon)
Add functions for interrogating the current
transaction log insertion point and determining
WAL filenames from
the hex WAL
locations displayed by
pg_stop_backup() and related functions.
Improve recovery from a crash during WAL replay (Simon)
The server now does periodic checkpoints during WAL recovery, so if there is a crash, future WAL recovery is shortened. This also eliminates the need for warm standby servers to replay the entire log since the base backup if they crash.
Improve reliability of long-term WAL replay (Heikki, Simon, Tom)
Formerly, trying to roll forward through more than 2 billion transactions would not work due to XID wraparound. This meant warm standby servers had to be reloaded from fresh base backups periodically.
archive_timeout to force transaction log
file switches at a given interval (Simon)
This enforces a maximum replication delay for warm standby servers.
Add native LDAP authentication (Magnus Hagander)
This is particularly useful for platforms that do not support PAM, such as Windows.
CONNECT ON DATABASE (Gevik Babakhani)
This gives SQL-level control over database access.
It works as an additional filter on top of the existing
Add support for SSL Certificate Revocation List (CRL) files (Libor Hohoš)
The server and libpq both recognize CRL files now.
GiST indexes are now clusterable (Teodor)
Remove routine autovacuum server log entries (Bruce)
pg_stat_activity now shows autovacuum
Track maximum XID age within individual tables, instead of whole databases (Alvaro)
This reduces the overhead involved in preventing transaction ID wraparound, by avoiding unnecessary VACUUMs.
Add last vacuum and analyze timestamp columns to the stats collector (Larry Rosenman)
These values now appear in the
pg_stat_*_tables system views.
Improve performance of statistics monitoring,
stats_command_string (Tom, Bruce)
This release enables
stats_command_string by default, now
that its overhead is minimal. This means
pg_stat_activity will now show all active
queries by default.
waiting column to
pg_stat_activity to show all the
information included in the ps display.
Add configuration parameter
update_process_title to control whether the
ps display is updated
for every command (Bruce)
On platforms where it is expensive to update the
ps display, it might
be worthwhile to turn this off and rely solely on
Allow units to be specified in configuration settings (Peter)
For example, you can now set
32MB rather than mentally converting
Add support for include directives
Improve logging of protocol-level prepare/bind/execute messages (Bruce, Tom)
Such logging now shows statement names, bind
parameter values, and the text of the query being
executed. Also, the query text is properly included in
logged error messages when enabled by
max_stack_depth from being set to unsafe
On platforms where we can determine the actual
kernel stack depth limit (which is most), make sure
that the initial default value of
max_stack_depth is safe, and reject
attempts to set it to unsafely large values.
Enable highlighting of error location in query in more cases (Tom)
The server is now able to report a specific error location for some semantic errors (such as unrecognized column name), rather than just for basic syntax errors as before.
to re-find parent key” errors in
pg_internal.init cache files during
server restart (Simon)
This avoids a hazard that the cache files might contain stale data after PITR recovery.
Fix race condition for truncation of a large
relation across a gigabyte boundary by
Fix bug causing needless deadlock errors on row-level locks (Tom)
Fix bugs affecting multi-gigabyte hash indexes (Tom)
Each backend process is now its own process group leader (Tom)
This allows query cancel to abort subprocesses invoked from a backend or archive/recovery process.
This allows these commands to return values, such as
the computed serial key for a new row. In the
UPDATE case, values from
the updated version of the row are returned.
Add support for multiple-row
VALUES clauses, per SQL standard
insert multiple rows of constants, or queries to
generate result sets using constants. For example,
INSERT ... VALUES (...), (...),
SELECT * FROM
(VALUES (...), (...), ....) AS alias(f1,
The SQL standard does not permit an alias in these commands, but many database systems allow one anyway for notational convenience.
UPDATE to set
multiple columns with a list of values (Susanne
This is basically a short-hand for assigning the
columns and values in pairs. The syntax is
UPDATE tab SET (.
column, ...) =
Make row comparisons work per standard (Tom)
The forms <, <=, >, >= now compare rows lexicographically, that is, compare the first elements, if equal compare the second elements, and so on. Formerly they expanded to an AND condition across all the elements, which was neither standard nor very useful.
automatically include all tables that reference the
specified table(s) via foreign keys. While convenient,
this is a dangerous tool — use with caution!
FOR UPDATE and
FOR SHARE in the same
IS NOT DISTINCT FROM (Pavel
This operator is similar to equality (
=), but evaluates to true when both
left and right operands are
NULL, and to false when just one is,
rather than yielding
in these cases.
Improve the length output used by
When all corresponding columns are of the same defined length, that length is used for the result, rather than a generic length.
ILIKE to work for multi-byte
lower() and then
regular expression patterns still do not work in these
standard_conforming_strings to be
This allows backslash escaping in strings to be
disabled, making PostgreSQL more
standards-compliant. The default is
off for backwards compatibility, but
future releases will default this to
Do not flatten subqueries that contain
volatile functions in their target
lists (Jaime Casanova)
This prevents surprising behavior due to multiple
evaluation of a
function (such as
nextval()). It might cause
performance degradation in the presence of functions
that are unnecessarily marked as
These are very useful in pooled connection setups.
This allows, for example, JDBC
parameters to work in these commands.
parameters are unspecified, infer their types from the
content of the query (Neil)
already did this.
OFFSET to exceed two
billion (Dhanaraj M)
This allows a tablespace to be specified for the new table.
ON COMMIT clause to
This allows temporary tables to be truncated or dropped on transaction commit. The default behavior is for the table to remain until the session ends.
LIKE (Greg Stark)
This allows easy copying of
CHECK constraints to a new table.
Allow the creation of placeholder (shell) types (Martijn van Oosterhout)
A shell type declaration creates a type name,
without specifying any of the details of the type.
Making a shell type is useful because it allows cleaner
declaration of the type's input/output functions, which
must exist before the type can be defined “for real”.
The syntax is
Aggregate functions now support multiple input parameters (Sergey Koposov, Tom)
Add new aggregate creation syntax (Tom)
The new syntax is
This more naturally supports the new multi-parameter
aggregate functionality. The previous syntax is still
ALTER ROLE PASSWORD
NULL to remove a previously set role
IF EXISTS for many object
operations on non-existent objects without generating
OWNED to drop all objects owned by a role
OWNED to reassign ownership of all objects
owned by a role (Alvaro)
above, facilitate dropping roles.
SEQUENCE syntax (Bruce)
This was added for setting sequence-specific
TABLE for sequences is still supported for
permission for sequences that allows only
USAGE permission allows
more fine-grained control over sequence access.
users to increment a sequence, but prevents them from
setting the sequence to an arbitrary value using
ALTER TABLE [
NO ] INHERIT (Greg Stark)
This allows inheritance to be adjusted dynamically, rather than just at table creation and destruction. This is very valuable when using inheritance to implement table partitioning.
Allow comments on global objects to be stored globally (Kris Jurka)
Previously, comments attached to databases were
stored in individual databases, making them
ineffective, and there was no provision at all for
comments on roles or tablespaces. This change adds a
new shared catalog
pg_shdescription and stores
comments on databases, roles, and tablespaces
Add option to allow indexes to be created without blocking concurrent writes to the table (Greg Stark, Tom)
The new syntax is
CONCURRENTLY. The default behavior is still
to block table modification while an index is being
Provide advisory locking functionality (Abhijit Menon-Sen, Tom)
This is a new locking API designed to replace what used to be in /contrib/userlock. The userlock code is now on pgfoundry.
COPY to dump a
SELECT query (Zoltan
Boszormenyi, Karel Zak)
dump arbitrary SQL
queries. The syntax is
(SELECT ...) TO.
return a command tag that includes the number of rows
copied (Volkan YAZICI)
expire rows without being affected by other concurrent
VACUUM operations (Hannu
Krossing, Alvaro, Tom)
detect the operating system locale and set the default
This makes it more likely that the installed
DateStyle value will be as
Reduce number of progress messages displayed by initdb (Tom)
Allow full timezone names in
timestamp input values (Joachim
Support configurable timezone abbreviations (Joachim Wieland)
A desired set of timezone abbreviations can be
chosen via the configuration parameter
the current wall-clock time,
statement_timestamp() is the time the
current statement arrived at the server, and
is an alias for
to_char() to print localized
month and day names (Euler Taveira de Oliveira)
Intervals and times are treated as 24-hour periods,
25 hours is
Add new function
justify_interval() to adjust interval units
Allow timezone offsets up to 14:59 away from GMT
Kiribati uses GMT+14, so we'd better accept that.
Interval computation improvements (Michael Glaesemann, Bruce)
Allow arrays to contain
NULL elements (Tom)
Allow assignment to array elements not contiguous with the existing entries (Tom)
The intervening array positions will be filled with nulls. This is per SQL standard.
New built-in operators
for array-subset comparisons (
&&) (Teodor, Tom)
These operators can be indexed for many data types using GiST or GIN indexes.
Add convenient arithmetic operations
CIDR values (Stephen R. van den Berg)
The new operators are
Add new aggregate functions from SQL:2003 (Neil)
The new functions are
stddev_samp() are merely renamings of
the existing aggregates
stddev(). The latter names remain
available for backward compatibility.
Add SQL:2003 statistical aggregates (Sergey Koposov)
Allow domains to be based on other domains (Tom)
Properly enforce domain
CHECK constraints everywhere
For example, the result of a user-defined function that is declared to return a domain type is now checked against the domain's constraints. This closes a significant hole in the domain implementation.
Fix problems with dumping renamed
SERIAL columns (Tom)
The fix is to dump a
SERIAL column by explicitly specifying
DEFAULT and sequence
elements, and reconstructing the
SERIAL column on reload using a new
OWNED BY command. This also allows dropping
Add a server-side sleep function
pg_sleep() (Joachim Wieland)
Add all comparison operators for the
tid (tuple id) data type (Mark
Kirkwood, Greg Stark, Tom)
TG_table_schema to trigger
TG_relname is now
deprecated. Comparable changes have been made in the
trigger parameters for the other PLs as well.
FOR statements to
return values to scalars as well as records and row
types (Pavel Stehule)
BY clause to the
FOR loop, to control the
iteration increment (Jaime Casanova)
SELECT INTO (Matt
STRICT mode throws an
exception if more or less than one row is returned by
table_schema to trigger
parameters (Adam Sjøgren)
Add prepared queries (Dmitry Karasik)
$_TD trigger data
a global variable (Andrew)
Previously, it was lexical, which caused unexpected sharing violations.
Run PL/Perl and PL/PerlU in separate interpreters, for security reasons (Andrew)
In consequence, they can no longer share data nor loaded modules. Also, if Perl has not been compiled with the requisite flags to allow multiple interpreters, only one of these languages can be used in any given backend process.
Named parameters are passed as ordinary variables,
as well as in the
array (Sven Suursoho)
table_schema to trigger
Allow returning of composite types and result sets (Sven Suursoho)
Return result-set as
generator (Sven Suursoho)
Allow functions to return
Python 2.5 is now supported (Tom)
Add new command
\password for changing role password
with client-side password encryption (Peter)
\c to connect to
a new host and port number (David, Volkan YAZICI)
Add tablespace display to
\l+ (Philip Yarra)
command to include the argument names and modes
INOUT) of the function (David
Add option to run the entire session in a single transaction (Simon)
Support for automatically retrieving
SELECT results in batches using a
cursor (Chris Mair)
This is enabled using
FETCH_COUNT . This feature
allows large result sets to be retrieved in
attempting to buffer the entire result set in
Make multi-line values align in the proper column (Martijn van Oosterhout)
Field values containing newlines are now displayed in a more readable fashion.
Save multi-line statements as a single entry, rather than one line at a time (Sergey E. Koposov)
This makes up-arrow recall of queries easier. (This is not available on Windows, because that platform uses the native command-line editing present in the operating system.)
Make the line counter 64-bit so it can handle files with more than two billion lines (David Fetter)
Report both the returned data and the command status
Allow complex selection of objects to be included or excluded by pg_dump (Greg Sabino Mullane)
options, and adds
-T options to exclude
objects. Also, the arguments of these switches can now
be wild-card expressions rather than single object
names, for example
'foo*', and a schema can be part of a
-T switch, for example
--no-data-for-failed-tables option to
suppress loading data if table creation failed (i.e.,
the table already exists) (Martin Pitt)
Add pg_restore option to run the entire session in a single transaction (Simon)
PQencryptPassword() to encrypt
This allows passwords to be sent pre-encrypted for
ALTER ROLE ...
This allows applications to query the thread-safety status of the library.
Allow a hostname in
~/.pgpass to match the default
socket directory (Bruce)
A blank hostname continues to match any Unix-socket connection, but this addition allows entries that are specific to one of several postmasters on the machine.
Allow MSVC to compile the PostgreSQL server (Magnus, Hiroshi Saito)
Add MSVC support for utility commands and pg_dump (Hiroshi Saito)
Add support for Windows code pages
1257 (Kris Jurka)
Drop privileges on startup, so that the server can be started from an administrative account (Magnus)
Stability fixes (Qingqing Zhou, Magnus)
Add native semaphore implementation (Qingqing Zhou)
The previous code mimicked SysV semaphores.
Add GIN (Generalized Inverted iNdex) index access method (Teodor, Oleg)
Remove R-tree indexing (Tom)
Rtree has been re-implemented using GiST. Among other differences, this means that rtree indexes now have support for crash recovery via write-ahead logging (WAL).
Reduce libraries needlessly linked into the backend (Martijn van Oosterhout, Tom)
Add a configure flag to allow libedit to be preferred over GNU readline (Bruce)
Allow installation into directories containing spaces (Peter)
Improve ability to relocate installation directories (Tom)
Add support for Solaris x86_64 using the Solaris compiler (Pierre Girard, Theo Schlossnagle, Bruce)
Add DTrace support (Robert Lor)
use by third-party applications wanting to test the
backend version in C using > and < comparisons
to report locking activity (Tom)
Emit warnings for unknown configure options (Martijn van Oosterhout)
Add server support for “plugin” libraries that can be used for add-on tasks such as debugging and performance measurement (Korry Douglas)
This consists of two features: a table of
variables” that allows separately-loaded
shared libraries to communicate, and a new
allows libraries to be loaded into specific sessions
without explicit cooperation from the client
application. This allows external add-ons to implement
features such as a PL/pgSQL debugger.
Rename existing configuration parameter
This was done for clarity in comparison to
Add new configuration parameter
(Greg Sabino Mullane)
This is like
server_version, but is an integer,
80200. This allows
applications to make version checks more easily.
Add a configuration parameter
Re-implement the regression test script as a C program (Magnus, Tom)
Allow loadable modules to allocate shared memory and lightweight locks (Marc Munro)
Add automatic initialization and finalization of dynamically loaded libraries (Ralf Engelschall, Tom)
_PG_fini() are called if
the library defines such symbols. Hence we no longer
need to specify an initialization function in
we can assume that the library used the
_PG_init() convention instead.
PG_MODULE_MAGIC header block to
all shared object files (Martijn van Oosterhout)
The magic block prevents version mismatches between loadable object files and servers.
Add shared library support for AIX (Laurenz Albe)
New XML documentation section (Bruce)
Major tsearch2 improvements (Oleg, Teodor)
multibyte encoding support, including UTF8
query rewriting support
improved ranking functions
thesaurus dictionary support
Ispell dictionaries now recognize MySpell format, used by OpenOffice
Add adminpack module containing Pgadmin administration functions (Dave)
These functions provide additional file system access routines not present in the default PostgreSQL server.
Add sslinfo module (Victor Wagner)
Reports information about the current connection's SSL certificate.
Add pgrowlocks module (Tatsuo)
This shows row locking information for a specified table.
Add hstore module (Oleg, Teodor)
Add isn module, replacing isbn_issn (Jeremy Kronuz)
This new implementation supports EAN13, UPC, ISBN (books), ISMN (music), and ISSN (serials).
Add index information functions to pgstattuple (ITAGAKI Takahiro, Satoshi Nagayasu)
Add pg_freespacemap module to display free space map information (Mark Kirkwood)
pgcrypto now has all planned functionality (Marko Kreen)
Include iMath library in pgcrypto to have the public-key encryption functions always available.
Add SHA224 algorithm that was missing in OpenBSD code.
Activate builtin code for SHA224/256/384/512 hashes on older OpenSSL to have those algorithms always available.
New function gen_random_bytes() that returns cryptographically strong randomness. Useful for generating encryption keys.
Remove digest_exists(), hmac_exists() and cipher_exists() functions.
Improvements to cube module (Joshua Reich)
New functions are
cube(float, float), and
Add async query capability to dblink (Kai Londenberg, Joe Conway)
New operators for array-subset comparisons
Various contrib packages already had these operators for their datatypes, but the naming wasn't consistent. We have now added consistently named array-subset comparison operators to the core code and all the contrib packages that have such functionality. (The old names remain available, but are deprecated.)
Add uninstall scripts for all contrib packages that have install scripts (David, Josh Drake)
If you see anything in the documentation that is not correct, does not match your experience with the particular feature or requires further clarification, please use this form to report a documentation issue.