Todo

From PostgreSQL wiki

(Difference between revisions)
Jump to: navigation, search
(ALTER: Update URLs for: |Reduce locking required for ALTER commands)
(XML: Done: Inline ORDER BY for XMLAGG. Example: "... XMLAGG(XMLELEMENT(...) ORDER BY col1) ...")
Line 582: Line 582:
 
|Extra functions from SQL:2006: XMLDOCUMENT, XMLCAST, XMLTEXT}}
 
|Extra functions from SQL:2006: XMLDOCUMENT, XMLCAST, XMLTEXT}}
  
{{TodoItem
+
{{TodoItemDone
 
|Inline ORDER BY for XMLAGG. Example: "... XMLAGG(XMLELEMENT(...) ORDER BY col1) ..." (should be made to work with all aggregate functions)}}
 
|Inline ORDER BY for XMLAGG. Example: "... XMLAGG(XMLELEMENT(...) ORDER BY col1) ..." (should be made to work with all aggregate functions)}}
  

Revision as of 21:00, 10 July 2010

Contents

This list contains all known PostgreSQL bugs and feature requests. If you would like to work on an item, please read the Developer FAQ first. There is also a development information page.

  • Incomplete item - marks ordinary, incomplete items
  • [E] Incomplete item - marks items that are easier to implement
  • [D] Completed item - marks changes that are done, and will appear in the next major release

For help on editing this list, please see Talk:Todo. Please do not add items here without discussion on the mailing list.

Administration

Incomplete itemAllow administrators to cancel multi-statement idle transactions

This allows locks to be released, but it is complex to report the cancellation back to the client.

Incomplete itemCheck for unreferenced table files created by transactions that were in-progress when the server terminated abruptly

Incomplete itemSet proper permissions on non-system schemas during db creation

Currently all schemas are owned by the super-user because they are copied from the template1 database. However, since all objects are inherited from the template database, it is not clear that setting schemas to the db owner is correct.

Incomplete itemAllow log_min_messages to be specified on a per-module basis

This would allow administrators to see more detailed information from specific sections of the backend, e.g. checkpoints, autovacuum, etc. Another idea is to allow separate configuration files for each module, or allow arbitrary SET commands to be passed to them. See also Logging Brainstorm.

Incomplete itemSimplify ability to create partitioned tables

This would allow creation of partitioned tables without requiring creation of triggers or rules for INSERT/UPDATE/DELETE, and constraints for rapid partition selection. Options could include range and hash partition selection. See also Table partitioning

Incomplete itemAllow auto-selection of partitioned tables for min/max() operations

There was a patch on -hackers from July 2009, but it has not been merged: MIN/MAX optimization for partitioned table

[D] Completed itemAllow more complex user/database default GUC settings

Currently ALTER USER and ALTER DATABASE support per-user and per-database defaults. Consider adding per-user-and-database defaults so things like search_path can be defaulted for a specific user connecting to a specific database.

Incomplete itemAllow custom variables to appear in pg_settings()

Incomplete itemHave custom variables be transaction safe

[D] Completed itemAllow custom variables to be created only by super-users but modified by ordinary uses, where appropriate

Incomplete itemImplement the SQL standard mechanism whereby REVOKE ROLE revokes only the privilege granted by the invoking role, and not those granted by other roles

Incomplete itemImprove server security options

Incomplete itemPrevent query cancel packets from being replayed by an attacker, especially when using SSL

Incomplete itemProvide a way to query the log collector subprocess to determine what the currently active log file is

Incomplete itemAllow the client to authenticate the server in a Unix-domain socket connection, e.g., using SO_PEERCRED

[D] Completed itemAllow the client to set an application_name to appear in pg_stat_activity

[D] Completed itemAllow server-side enforcement of password policies

Password checks might include password complexity or non-reuse of passwords. This facility will require the client to send password creation/changes to the server in plain-text, not MD5.

Incomplete itemAllow custom daemons to be automatically stopped/started along postmaster

This allows easier administration of daemons like user job schedulers or replication-related daemons.

Configuration files

Incomplete itemAllow pg_hba.conf to specify host names along with IP addresses

Host name lookup could occur when the postmaster reads the pg_hba.conf file, or when the backend starts. Another solution would be to reverse lookup the connection IP and check that hostname against the host names in pg_hba.conf. We could also then check that the host name maps to the IP address.

Incomplete itemAllow postgresql.conf file values to be changed via an SQL API, perhaps using SET GLOBAL

Incomplete itemAllow the server to be stopped/restarted via an SQL API

Incomplete itemConsider normalizing fractions in postgresql.conf, perhaps using '%'

Incomplete itemAllow Kerberos to disable stripping of realms so we can check the username@realm against multiple realms

Incomplete itemAdd functions to check correctness of configuration files before they are loaded "live"

Incomplete itemImprove LDAP authentication configuration options

Incomplete itemAdd external tool to auto-tune some postgresql.conf parameters

Incomplete itemAdd 'hostgss' pg_hba.conf option to allow GSS link-level encryption

Incomplete itemProcess pg_hba.conf keywords as case-insensitive

Tablespaces

Incomplete itemAllow a database in tablespace t1 with tables created in tablespace t2 to be used as a template for a new database created with default tablespace t2

Currently all objects in the default database tablespace must have default tablespace specifications. This is because new databases are created by copying directories. If you mix default tablespace tables and tablespace-specified tables in the same directory, creating a new database from such a mixed directory would create a new database with tables that had incorrect explicit tablespaces. To fix this would require modifying pg_class in the newly copied database, which we don't currently do.

Incomplete itemAllow reporting of which objects are in which tablespaces

This item is difficult because a tablespace can contain objects from multiple databases. There is a server-side function that returns the databases which use a specific tablespace, so this requires a tool that will call that function and connect to each database to find the objects in each database for that tablespace.

Incomplete itemAllow WAL replay of CREATE TABLESPACE to work when the directory structure on the recovery computer is different from the original

Incomplete itemAllow per-tablespace quotas

[D] Completed itemAllow per-tablespace random_page_cost

Statistics Collector

Incomplete itemAllow statistics last vacuum/analyze execution times to be displayed without requiring track_counts to be enabled

Incomplete itemClear table counters on TRUNCATE

Incomplete item Allow the clearing of cluster-level statistics

Point-In-Time Recovery (PITR)

[D] Completed itemAllow a warm standby system to also allow read-only statements

[E] Incomplete itemCreate dump tool for write-ahead logs for use in determining transaction id for point-in-time recovery

This is useful for checking PITR recovery.

Incomplete itemAllow recovery.conf to support the same syntax as postgresql.conf, including quoting

Incomplete itemAllow archive_mode to be changed without server restart?

Incomplete itemConsider avoiding WAL switching via archive_timeout if there has been no database activity

[E] Incomplete itemExpose pg_controldata via SQL interface

Helpful for monitoring replicated databases; initial patch

SSL

Incomplete itemAllow SSL authentication/encryption over unix domain sockets

Incomplete itemAllow SSL key file permission checks to be optionally disabled when sharing SSL keys with other applications

Incomplete itemAllow SSL CRL files to be re-read during configuration file reload, rather than requiring a server restart

Unlike SSL CRT files, CRL (Certificate Revocation List) files are updated frequently

Alternatively or additionally supporting OCSP (online certificate security protocol) would provide real-time revocation discovery without reloading

Incomplete item Allow automatic selection of SSL client certificates from a certificate store

Incomplete item Send full certificate server chain to client

Standby server mode

Incomplete item Allow pg_xlogfile_name() to be used in recovery mode

[D] Completed item Redefine smart shutdown in standby mode to exist as soon as all read-only connections are gone.

Incomplete item Fix things so that any such variables inherited from the server environment are intentionally *NOT* used for making SR connections.

[D] Completed item Consider changing the calculation of superused_reserved_connections so that replication connections don't consume superuser_reserved_connections slots.

Incomplete item Add a new privilege for connecting for streaming replication

Incomplete item Add support for synchronous replication.

Incomplete item Add capability to take and send a base backup over the streaming replication connection, making it possible to initialize a new standby server from a running primary server without a WAL archive or other access to the primary server.


Data Types

Incomplete itemChange NUMERIC to enforce the maximum precision

Incomplete itemReduce storage space for small NUMERICs

Incomplete itemFix data types where equality comparison isn't intuitive, e.g. box

Incomplete itemAdd support for public SYNONYMs

Incomplete itemAdd support for SQL-standard GENERATED/IDENTITY columns

Incomplete itemConsider placing all sequences in a single table, or create a system view

Incomplete itemConsider a special data type for regular expressions

Incomplete itemReduce BIT data type overhead using short varlena headers

Incomplete itemAllow adding/renaming/removing enumerated values to an existing enumerated data type

Incomplete itemSupport scoped IPv6 addresses in inet type

Incomplete itemAdd JSON (JavaScript Object Notation) data type

This would behave similar to the XML data type, which is stored as text, but allows element lookup and conversion functions.


Incomplete itemConsidering improving performance of computing CHAR() value lengths

Domains

Dates and Times

Incomplete itemAllow infinite intervals just like infinite timestamps

Incomplete itemAllow TIMESTAMP WITH TIME ZONE to store the original timezone information, either zone name or offset from UTC

If the TIMESTAMP value is stored with a time zone name, interval computations should adjust based on the time zone rules.

Incomplete itemFix SELECT '0.01 years'::interval, '0.01 months'::interval

Incomplete itemHave timestamp subtraction not call justify_hours()?

Incomplete itemImprove timestamptz subtraction to be DST-aware

Currently subtracting one date from another that crosses a daylight savings time adjustment can return '1 day 1 hour', but adding that back to the first date returns a time one hour in the future. This is caused by the adjustment of '25 hours' to '1 day 1 hour', and '1 day' is the same time the next day, even if daylight savings adjustments are involved.

Incomplete itemFix interval display to support values exceeding 2^31 hours

Incomplete itemAdd overflow checking to timestamp and interval arithmetic

[D] Completed itemRevise the src/timezone/tznames abbreviation files:

Incomplete itemAdd function to allow the creation of timestamps using parameters

Arrays

Incomplete itemAdd support for arrays of domains

Incomplete itemAllow single-byte header storage for array elements

Incomplete itemAdd function to detect if an array is empty

Incomplete itemImprove handling of empty arrays

Incomplete itemImprove handling of NULLs in arrays

Binary Data

Incomplete itemImprove vacuum of large objects, like contrib/vacuumlo?

[D] Completed itemAdd security checks for large objects

Incomplete itemAuto-delete large objects when referencing row is deleted

contrib/lo offers this functionality.

Incomplete itemAllow read/write into TOAST values like large objects

This requires the TOAST column to be stored EXTERNAL.

Incomplete itemAdd API for 64-bit large object access

MONEY Data Type

Incomplete itemAdd locale-aware MONEY type, and support multiple currencies

Incomplete itemMONEY dumps in a locale-specific format making it difficult to restore to a system with a different locale

Incomplete itemAllow MONEY to be easily cast to/from other numeric data types

Text Search

Incomplete itemAllow dictionaries to change the token that is passed on to later dictionaries

Incomplete itemConsider a function-based API for '@@' searches

Incomplete itemImprove text search error messages

Incomplete itemConsider changing error to warning for strings larger than one megabyte

Incomplete itemtsearch and tsdicts regression tests fail in Turkish locale on glibc

Incomplete itemtsquery negator operator treated as part of lexeme

XML

Incomplete itemAllow xml arrays to be cast to other data types

Incomplete itemAdd XML Schema validation and xmlvalidate function (SQL:2008)

Incomplete itemAdd xmlvalidatedtd variant to support validating against a DTD?

Incomplete itemRelax-NG validation; libxml2 supports this already

Incomplete itemMake it work reliably for non-UTF8 server encoding (xpath()) in particular is known to not work)

Incomplete itemExtra functions from SQL:2006: XMLDOCUMENT, XMLCAST, XMLTEXT

[D] Completed itemInline ORDER BY for XMLAGG. Example: "... XMLAGG(XMLELEMENT(...) ORDER BY col1) ..." (should be made to work with all aggregate functions)

Incomplete itemXMLNAMESPACES support in XMLELEMENT and elsewhere

Incomplete itemXSLT support; already available in contrib/xml2, but needs API fixes and adaptation to xml type.

Incomplete itemXML Canonical: Convert XML documents to canonical form to compare them. libxml2 has support for this.

Incomplete itemPretty-printing XML: Parse a document and serialize it back in some indented form. libxml2 might support this.

Incomplete itemXMLQUERY (from SQL/XML standard)

Incomplete itemIn some cases shredding could be better option (if there is no need in keeping XML docs entirely; if we have already developed tools that understand only relational data; etc) -- it would be a separate module that implements annotated schema decomposition technique, similar to DB2 and SQL Server functionality.

Incomplete item Nested or repeated xpath() apparently mess up namespaces [1] [2] [3] [4] [5]

Incomplete itemXPath: Adding the <x> at the root causes problems [6] [7] [8]

Incomplete itemxpath_table needs to be implemented/implementable to get rid of contrib/xml2 [9]

Incomplete itemxpath_table is pretty broken anyway [10]

Incomplete itembetter handling of XPath data types [11] [12]

Incomplete itemxpath_exists() is needed. It checks, whether or not the path specified exists in the XML value. (W/o this function we need to use weird "array_dims(xpath(...)) IS NOT NULL" syntax.)

Incomplete itembetter handling of PIs and DTDs in xmlconcat() [13]

Functions

Incomplete itemAllow INET subnet tests using non-constants to be indexed

Incomplete itemAdd INET overlaps operator, for use by exclusion constraints

Incomplete itemAllow to_date() and to_timestamp() to accept localized month names

Incomplete itemAdd missing parameter handling in to_char()

Incomplete itemThrow an error from to_char() instead of printing a string of "#" when a number doesn't fit in the desired output format.

[D] Completed itemAdd functions to get/set bit values

Incomplete itemAllow to_char() on interval values to accumulate the highest unit requested

Some special format flag would be required to request such accumulation. Such functionality could also be added to EXTRACT. Prevent accumulation that crosses the month/day boundary because of the uneven number of days in a month.

  • to_char(INTERVAL '1 hour 5 minutes', 'MI') => 65
  • to_char(INTERVAL '43 hours 20 minutes', 'MI' ) => 2600
  • to_char(INTERVAL '43 hours 20 minutes', 'WK:DD:HR:MI') => 0:1:19:20
  • to_char(INTERVAL '3 years 5 months','MM') => 41

Incomplete itemAllow SQL-language functions to reference parameters by parameter name

Currently SQL-language functions can only refer to dollar parameters, e.g. $1

Incomplete itemAdd SPI_gettypmod() to return the typemod for a TupleDesc

Incomplete itemEnforce typmod for function inputs, function results and parameters for spi_prepare'd statements called from PLs

Incomplete itemAllow holdable cursors in SPI

Incomplete itemTighten function permission checks

Incomplete itemFix IS OF so it matches the ISO specification, and add documentation

[D] Completed itemAdd missing operators for geometric data types

Some geometric types do not have the full suite of geometric operators, e.g. box @> point

Incomplete itemAdd overlaps geometric operators that ignore point overlaps

Incomplete itemImplement Boyer-Moore searching in LIKE queries

Incomplete itemPrevent malicious functions from being executed with the permissions of unsuspecting users

Index functions are safe, so VACUUM and ANALYZE are safe too. Triggers, CHECK and DEFAULT expressions, and rules are still vulnerable.

Incomplete itemReduce memory usage of aggregates in set returning functions

Incomplete itemFix /contrib/ltree operator

Incomplete itemFix inconsistent precedence of =, >, and < compared to <>, >=, and <=

Incomplete itemFix regular expression bug when using complex back-references

Incomplete itemHave /contrib/dblink reuse unnamed connections

Incomplete itemAllow calling of a procedure outside a SELECT that can control the transaction state

[D] Completed itemAdd has_sequence_privilege()

Incomplete itemImprove formatting of pg_get_viewdef() output

Incomplete itemAdd printf()-like functionality

Incomplete itemFix to_number() handling for values not matching the format string

Incomplete itemAdd function to dump pg_depend information cleanly

Multi-Language Support

Incomplete itemAdd NCHAR (as distinguished from ordinary varchar),

Incomplete itemAllow more fine-grained collation selection; add CREATE COLLATION.

Right now the collation is fixed at database creation time.

Incomplete itemAdd a LOCALE option to CREATE DATABASE, as a shorthand

Incomplete itemSupport multiple simultaneous character sets, per SQL:2008

Incomplete itemImprove UTF8 combined character handling?

Incomplete itemAdd octet_length_server() and octet_length_client()

Incomplete itemMake octet_length_client() the same as octet_length()?

Incomplete itemFix problems with wrong runtime encoding conversion for NLS message files

Incomplete itemAdd URL to more complete multi-byte regression tests

Incomplete itemFix contrib/fuzzystrmatch to work with multibyte encodings

Incomplete itemSet client encoding based on the client operating system encoding

Currently client_encoding is set in postgresql.conf, which defaults to the server encoding.

Incomplete itemChange memory allocation for multi-byte functions so memory is allocated inside conversion functions

Currently we preallocate memory based on worst-case usage.

Incomplete itemAdd ability to use case-insensitive regular expressions on multi-byte characters

ILIKE already works with multi-byte characters

Incomplete itemImprove encoding of connection startup messages sent to the client

Currently some authentication error messages are sent in the server encoding

Incomplete itemHave pg_stat_activity display query strings in the correct client encoding

Incomplete itemMore sensible support for Unicode combining characters, normal forms

Views / Rules

Incomplete itemAutomatically create rules on views so they are updateable, per SQL:2008

We can only auto-create rules for simple views. For more complex cases users will still have to write rules manually.

Incomplete itemAdd the functionality for WITH CHECK OPTION clause of CREATE VIEW

Incomplete itemAllow VIEW/RULE recompilation when the underlying tables change

This is both difficult and controversial.

Incomplete itemMake it possible to use RETURNING together with conditional DO INSTEAD rules, such as for partitioning setups

Incomplete itemAdd the ability to automatically create materialized views

Right now materialized views require the user to create triggers on the main table to keep the summary table current. SQL syntax should be able to manage the triggers and summary table automatically. A more sophisticated implementation would automatically retrieve from the summary table when the main table is referenced, if possible. See Materialized Views for implementation details

Incomplete itemImprove ability to modify views via ALTER TABLE

Incomplete itemPrevent low-cost functions from seeing unauthorized view rows

SQL Commands

Incomplete itemAdd CORRESPONDING BY to UNION/INTERSECT/EXCEPT

Incomplete itemAdd ROLLUP, CUBE, GROUPING SETS options to GROUP BY

[E] Incomplete itemAllow SET CONSTRAINTS to be qualified by schema/table name

[E] Incomplete itemFix TRUNCATE ... RESTART IDENTITY so its effect on sequences is rolled back on transaction abort

Incomplete itemAllow PREPARE of cursors

Incomplete itemAllow finer control over the caching of prepared query plans

Currently anonymous (un-named) queries prepared via the wire protocol are replanned every time bind parameters are supplied --- allow SQL PREPARE to do the same. Also, allow control over replanning prepared queries either manually or automatically when statistics for execute parameters differ dramatically from those used during planning.

Incomplete itemImprove logging of prepared transactions recovered during startup

Incomplete itemAllow prepared transactions with temporary tables created and dropped in the same transaction, and when an ON COMMIT DELETE ROWS temporary table is accessed

Incomplete itemAdd a GUC variable to warn about non-standard SQL usage in queries

Incomplete itemAdd SQL-standard MERGE/REPLACE/UPSERT command

MERGE is typically used to merge two tables. REPLACE or UPSERT command does UPDATE, or on failure, INSERT. See SQL MERGE for notes on the implementation details.

Incomplete itemAdd NOVICE output level for helpful messages like automatic sequence/index creation

Incomplete itemAdd GUC to issue notice about statements that use unjoined tables

Incomplete itemAllow EXPLAIN to identify tables that were skipped because of constraint_exclusion

[D] Completed itemAllow EXPLAIN output to be more easily processed by scripts, perhaps XML

Incomplete itemEnable standard_conforming_strings by default in 9.1?

When this is done, backslash-quote should be prohibited in non-E'' strings because of possible confusion over how such strings treat backslashes. Basically, '' is always safe for a literal single quote, while \' might or might not be based on the backslash handling rules.

Incomplete itemSimplify dropping roles that have objects in several databases

Incomplete itemAllow the count returned by SELECT, etc to be represented as an int64 to allow a higher range of values

Incomplete itemAdd support for WITH RECURSIVE ... CYCLE

Incomplete itemAdd DEFAULT .. AS OWNER so permission checks are done as the table owner

This would be useful for SERIAL nextval() calls and CHECK constraints.

Incomplete itemAllow DISTINCT to work in multiple-argument aggregate calls

Incomplete itemAdd column to pg_stat_activity that shows the progress of long-running commands like CREATE INDEX and VACUUM

Incomplete itemAllow INSERT/UPDATE/DELETE ... RETURNING inside a SELECT 'FROM' clause or target list

Actually it would be saner to allow this in WITH

Incomplete itemAllow INSERT/UPDATE/DELETE ... RETURNING in common table expressions

Incomplete itemAdd comments on system tables/columns using the information in catalogs.sgml

Ideally the information would be pulled from the SGML file automatically.

Incomplete itemPrevent the specification of conflicting transaction read/write options

Incomplete itemSupport LATERAL subqueries

Lateral subqueries can reference columns of tables defined outside the subquery at the same level, i.e. laterally. For example, a LATERAL subquery in a FROM clause could reference tables defined in the same FROM clause. Currently only the columns of tables defined above subqueries are recognized.

[D] Completed itemForbid COMMENT on columns of an index

Postgres currently allows comments to be placed on the columns of an index, but pg_dump doesn't handle them and the column names themselves are implementation-dependent.

Incomplete itemAdd support for functional dependencies

This would allow omitting GROUP BY columns when grouping by the primary key.

Incomplete itemOptimize ON COMMIT DELETE ROWS

Currently deletions are excessively performed

CREATE

Incomplete itemAllow CREATE TABLE AS to determine column lengths for complex expressions like SELECT col1

Incomplete itemHave WITH CONSTRAINTS also create constraint indexes

Incomplete itemMove NOT NULL constraint information to pg_constraint

Currently NOT NULL constraints are stored in pg_attribute without any designation of their origins, e.g. primary keys. One manifest problem is that dropping a PRIMARY KEY constraint does not remove the NOT NULL constraint designation. Another issue is that we should probably force NOT NULL to be propagated from parent tables to children, just as CHECK constraints are. (But then does dropping PRIMARY KEY affect children?)

Incomplete itemPrevent concurrent CREATE TABLE from sometimes returning a cryptic error message

[D] Completed itemAdd CREATE TABLE LIKE ... INCLUDING COMMENTS

Incomplete itemAllow CREATE TABLE to optionally create a table if it does not already exist, without throwing an error

The fact that tables contain data makes this more complex than other CREATE OR REPLACE operations.

Incomplete itemAdd CREATE SCHEMA ... LIKE that copies a schema

[D] Completed itemHave CREATE TABLE LIKE copy column storage parameters

Incomplete itemCREATE OR REPLACE FUNCTION might leave dependent objects depending on the function in inconsistent state

Incomplete itemAllow temporary tables to exist as empty by default in all sessions

[D] Completed itemAdd OR REPLACE to CREATE LANGUAGE

Incomplete itemAllow the creation of "distinct" types

Incomplete itemConsider analyzing temporary tables when they are first used in a query

Autovacuum cannot analyze or vacuum temporary tables.

UPDATE

Incomplete itemAllow UPDATE tab SET ROW (col, ...) = (SELECT...)

Incomplete itemResearch self-referential UPDATEs that see inconsistent row versions in read-committed mode

Incomplete itemImprove performance of EvalPlanQual mechanism that rechecks already-updated rows

This is related to the previous item, which questions whether it even has the right semantics

ALTER

Incomplete itemHave ALTER TABLE RENAME rename SERIAL sequence names

[E] Incomplete itemAllow ALTER TABLE ... ALTER CONSTRAINT ... RENAME

Incomplete itemAdd ALTER TABLE RENAME CONSTRAINT

Incomplete itemHave ALTER SEQUENCE RENAME rename the sequence name stored in the sequence table

Incomplete itemAdd ALTER DOMAIN to modify the underlying data type

[E] Incomplete itemAllow ALTER TABLE to change constraint deferrability and actions

Incomplete itemAdd missing object types for ALTER ... SET SCHEMA

Incomplete itemAllow ALTER TABLESPACE to move to different directories

Incomplete itemAllow moving system tables to other tablespaces, where possible

Currently non-global system tables must be in the default database tablespace. Global system tables can never be moved.

Incomplete itemHave ALTER INDEX update the name of a constraint using that index

Incomplete itemAllow column display reordering by recording a display, storage, and permanent id for every column?

Incomplete itemAllow an existing index to be marked as a table's primary key

Incomplete itemAllow ALTER TYPE on composite types to perform operations similar to ALTER TABLE

Incomplete itemDon't require table rewrite on ALTER TABLE ... ALTER COLUMN TYPE, when the old and new data types are binary compatible

Incomplete itemReduce locking required for ALTER commands

CLUSTER

Incomplete itemAutomatically maintain clustering on a table

This might require some background daemon to maintain clustering during periods of low usage. It might also require tables to be only partially filled for easier reorganization. Another idea would be to create a merged heap/index data file so an index lookup would automatically access the heap data too. A third idea would be to store heap rows in hashed groups, perhaps using a user-supplied hash function.

[E] Incomplete itemAdd default clustering to system tables

To do this, determine the ideal cluster index for each system table and set the cluster setting during initdb.

Incomplete itemImprove CLUSTER performance by sorting to reduce random I/O

[E] Incomplete itemMake CLUSTER VERBOSE more verbose.

It is also used by new VACUUM FULL VERBOSE.

COPY

Incomplete itemAllow COPY to report error lines and continue

This requires the use of a savepoint before each COPY line is processed, with ROLLBACK on COPY failure.

Incomplete itemAllow COPY on a newly-created table to skip WAL logging

On crash recovery, the table involved in the COPY would be removed or have its heap and index files truncated. One issue is that no other backend should be able to add to the table at the same time, which is something that is currently allowed. This currently is done if the table is created inside the same transaction block as the COPY because no other backends can see the table.

Incomplete itemAllow COPY FROM to create index entries in bulk

Incomplete itemAllow COPY in CSV mode to control whether a quoted zero-length string is treated as NULL

Currently this is always treated as a zero-length string, which generates an error when loading into an integer column

Incomplete itemImprove COPY performance

Incomplete itemAllow COPY to report errors sooner

[D] Completed itemImprove bytea COPY format

Incomplete itemAllow COPY to handle other number formats eg. the German notation. Best would be something like WITH DECIMAL ','.

Incomplete itemAllow a stalled COPY to exit if the backend is terminated

GRANT/REVOKE

[D] Completed itemAllow GRANT/REVOKE permissions to be applied to all schema objects with one command

The proposed syntax is: GRANT SELECT ON ALL TABLES IN public TO phpuser; GRANT SELECT ON NEW TABLES IN public TO phpuser;

[D] Completed itemAllow GRANT/REVOKE permissions to be inherited by objects based on schema permissions

Incomplete itemAllow SERIAL sequences to inherit permissions from the base table?

Incomplete itemAllow dropping of a role that has connection rights

DECLARE CURSOR

Incomplete itemPrevent DROP TABLE from dropping a table referenced by its own open cursor?

Incomplete itemProvide some guarantees about the behavior of cursors that invoke volatile functions

INSERT

Incomplete itemAllow INSERT/UPDATE of the system-generated oid value for a row

Incomplete itemIn rules, allow VALUES() to contain a mixture of 'old' and 'new' references

SHOW/SET

Incomplete itemAdd SET PERFORMANCE_TIPS option to suggest INDEX, VACUUM, VACUUM ANALYZE, and CLUSTER

Incomplete itemRationalize the discrepancy between settings that use values in bytes and SHOW that returns the object count

LISTEN/NOTIFY

[D] Completed itemAllow LISTEN/NOTIFY to store info in memory rather than tables

Currently LISTEN/NOTIFY information is stored in pg_listener. Storing such information in memory would improve performance.

[D] Completed itemAdd optional textual message to NOTIFY

This would allow an informational message to be added to the notify message, perhaps indicating the row modified or other custom information.

Incomplete itemAllow NOTIFY in rules involving conditionals

Window Functions

See TODO items for window functions.

Incomplete itemSupport creation of user-defined window functions.

We have the ability to create new window functions written in C. Is it worth the effort to create an API that would let them be written in PL/pgsql, etc?

Incomplete itemImplement full support for window framing clauses.

In addition to done clauses described in the latest doc, these clauses are not implemented yet.

  • RANGE BETWEEN ... PRECEDING/FOLLOWING
  • EXCLUDE

Incomplete itemLook at tuplestore performance issues.

The tuplestore_in_memory() thing is just a band-aid, we ought to try to solve it properly. tuplestore_advance seems like a weak spot as well.

Incomplete itemDo we really need so much duplicated code between Agg and WindowAgg?

Incomplete itemTeach planner to evaluate multiple windows in the optimal order.

Currently windows are always evaluated in the query-specified order.

Incomplete itemImplement DISTINCT clause in window aggregates.

Some proprietary RDBMSs have implemented it already, so it helps with porting from those.

Integrity Constraints

Keys

[D] Completed itemAllow DEFERRABLE UNIQUE constraints

This would allow UPDATE tab SET col = col + 1 to work if col has a unique index. Currently, uniqueness checks are done while the command is being executed, rather than at the end of the statement or transaction.

Incomplete itemImprove deferrable unique constraints for cases with many conflicts

The current implementation fires a trigger for each potentially conflicting row. This might not scale well for an update that changes many key values at once.

Referential Integrity

Incomplete itemAdd MATCH PARTIAL referential integrity

Incomplete itemChange foreign key constraint for array -> element to mean element in array?

Incomplete itemFix problem when cascading referential triggers make changes on cascaded tables, seeing the tables in an intermediate state

Incomplete itemOptimize referential integrity checks

Server-Side Languages

Incomplete itemAdd support for polymorphic arguments and return types to languages other than PL/PgSQL

Incomplete itemAdd capability to create and call PROCEDURES

Incomplete itemAdd support for OUT and INOUT parameters to languages other than PL/PgSQL

Incomplete itemAdd more fine-grained specification of functions taking arbitrary data types

Incomplete itemAdd support for returning multiple result sets?

PL/pgSQL

[D] Completed itemAllow function parameters to be passed by name, get_employee_salary(12345 AS emp_id, 2001 AS tax_year)

Incomplete itemAllow handling of %TYPE arrays, e.g. tab.col%TYPE[]

Incomplete itemAllow listing of record column names, and access to record columns via variables, e.g. columns := r.(*), tval2 := r.(colname)

Incomplete itemAdd support for SCROLL cursors

Incomplete itemAdd support for WITH HOLD cursors

Incomplete itemAllow row and record variables to be set to NULL constants, and allow NULL tests on such variables

Because a row is not scalar, do not allow assignment from NULL-valued scalars.

[D] Completed itemReview handling of MOVE and FETCH

[D] Completed itemImprove logic of determining if an identifier is a variable or a column name

Incomplete itemConsider keeping separate cached copies when search_path changes

[D] Completed itemImprove PL/pgSQL's ability to cope with rowtypes containing dropped columns

Incomplete itemImprove handling of NULL row values vs. NULL rows

PL/Perl

Incomplete itemAllow data to be passed in native language formats, rather than only text

Incomplete itemAllow regex operations in plperl using UTF8 characters in non-UTF8 encoded databases.

PL/Python

Incomplete itemAdd table function support

Incomplete itemAdd tracebacks

[D] Completed itemAdd support for Python 3

Incomplete itemDevelop a trusted variant of PL/Python.

Incomplete itemAllow arrays as function arguments and return values.

Incomplete itemCreate a new restricted execution class that will allow passing function arguments in as locals. Passing them as globals means functions cannot be called recursively.

Incomplete itemFunctions cache the input and output functions for their arguments, so the following will make PostgreSQL unhappy:

create table users (first_name text, last_name text);
create function user_name(user) returns text as 'mycode' language plpython;
select user_name(user) from users;
alter table add column user_id integer;
select user_name(user) from users;

You have to drop and create the function(s) each time its arguments are modified (not nice), or don't cache the input and output functions (slower?), or check if the structure of the argument has been altered (is this possible, easy, quick?) and recreate cache.

Incomplete itemBetter documentation

Incomplete itemAdd a DB-API compliant interface on top of the SPI interface.

Incomplete itemcheck encoding validity of values passed back to Postgres in function returns, trigger tuple changes, or SPI calls.

PL/Tcl

Incomplete itemAdd table function support

Incomplete itemcheck encoding validity of values passed back to Postgres in function returns, trigger tuple changes, or SPI calls.

Clients

Incomplete itemAdd a function like pg_get_indexdef() that report more detailed index information

pg_ctl

Incomplete itemAllow pg_ctl to work properly with configuration files located outside the PGDATA directory

pg_ctl can not read the pid file because it isn't located in the config directory but in the PGDATA directory. The solution is to allow pg_ctl to read and understand postgresql.conf to find the data_directory value.

Incomplete itemHave the postmaster write a random number to a file on startup that pg_ctl checks against the contents of a pg_ping response on its initial connection (without login)

This will protect against connecting to an old instance of the postmaster in a different or deleted subdirectory.

Incomplete itemModify pg_ctl behavior and exit codes to make it easier to write an LSB conforming init script

It may be desirable to condition some of the changes on a command-line switch, to avoid breaking existing scripts. A Linux shell (sh) script is referenced which has been tested and seems to provide a high degree of conformance in multiple environments. Study of this script might suggest areas where pg_ctl could be modified to make writing an LSB conforming script easier; however, some aspects of that script would be unnecessary with other suggested changes to pg_ctl, and discussion on the lists did not reach consensus on support for all aspects of this script. Further discussion of particular changes is needed before beginning any work.

These threads should be studied for other ideas on improvements:


psql

Incomplete itemHave psql \ds show all sequences and their settings

Incomplete itemHave \d on a sequence indicate if the sequences is owned by a table

Incomplete itemMove psql backslash database information into the backend, use mnemonic commands?

This would allow non-psql clients to pull the same information out of the database as psql.

Incomplete itemMake psql's \d commands more consistent in its handling of schemas

Incomplete itemConsistently display privilege information for all objects in psql

Incomplete itemAdd auto-expanded mode so expanded output is used if the row length is wider than the screen width.

Consider using auto-expanded mode for backslash commands like \df+.

Incomplete itemPrevent tab completion of SET TRANSACTION from querying the database and therefore preventing the transaction isolation level from being set.

Currently SET <tab> causes a database lookup to check all supported session variables. This query causes problems because setting the transaction isolation level must be the first statement of a transaction.

Incomplete itemAdd a \set variable to control whether \s displays line numbers

Another option is to add \# which lists line numbers, and allows command execution.

Incomplete itemPrevent escape string warnings when object names have backslashes

Incomplete itemHave \d show child tables that inherit from the specified parent

Incomplete itemInclude the symbolic SQLSTATE name in verbose error reports

Incomplete itemAdd prompt escape to display the client and server versions

Incomplete itemAdd option to wrap column values at whitespace boundaries, rather than chopping them at a fixed width.

Currently, "wrapped" format chops values into fixed widths. Perhaps the word wrapping could use the same algorithm documented in the W3C specification.

Incomplete itemAdd "auto" expanded mode that outputs in expanded format if "wrapped" mode can't wrap the output to the screen width

Incomplete itemSupport the ReST table output format

Details about the ReST format: http://docutils.sourceforge.net/rst.html#reference-documentation

Incomplete itemAdd option to print advice for people familiar with other databases

Incomplete itemConsider showing TOAST and index sizes in \dt+

Incomplete itemAllow \dd to show constraint comments

Incomplete itemAdd ability to edit views with \ev

Incomplete itemAdd \dL to show languages

[E] Incomplete itemDistinguish between unique indexes and unique constraints in \d+

Incomplete itemFix FETCH_COUNT to handle SELECT ... INTO and WITH queries

Incomplete itemPrevent psql from sending remaining single-line multi-statement queries after reconnecting

pg_dump / pg_restore

[D] Completed itemAdd dumping of comments on composite type columns

[E] Incomplete itemAdd full object name to the tag field. eg. for operators we need '=(integer, integer)', instead of just '='.

[D] Completed item Add comments to output indicating version of pg_dump and of the database server

Incomplete itemAdd pg_dumpall custom format dumps?

Incomplete itemAvoid using platform-dependent locale names in pg_dumpall output

Using native locale names puts roadblocks in the way of porting a dump to another platform. One possible solution is to get CREATE DATABASE to accept some agreed-on set of locale names and fix them up to meet the platform's requirements.

Incomplete itemAllow selection of individual object(s) of all types, not just tables

Incomplete itemIn a selective dump, allow dumping of an object and all its dependencies

Incomplete itemAdd options like pg_restore -l and -L to pg_dump

Incomplete itemAdd support for multiple pg_restore -t options, like pg_dump

pg_restore's -t switch is less useful than pg_dump's in quite a few ways: no multiple switches, no pattern matching, no ability to pick up indexes and other dependent items for a selected table. It should be made to handle this switch just like pg_dump does.

Incomplete itemStop dumping CASCADE on DROP TYPE commands in clean mode

Incomplete itemAllow pg_dump --clean to drop roles that own objects or have privileges

tgl says: if this is about pg_dumpall, it's done as of 8.4. If it's really about pg_dump, what does it mean? pg_dump has no business dropping roles.

Incomplete itemRemove unnecessary function pointer abstractions in pg_dump source code

Incomplete itemAllow pg_dump to utilize multiple CPUs and I/O channels by dumping multiple objects simultaneously

The difficulty with this is getting multiple dump processes to produce a single dump output file. It also would require several sessions to share the same snapshot.

Incomplete itemAllow pg_restore to load different parts of the COPY data for a single table simultaneously

Incomplete itemRemove support for dumping from pre-7.3 servers

In 7.3 and later, we can get accurate dependency information from the server. pg_dump still contains a lot of crufty code to try to deal with the lack of dependency info in older servers, but the usefulness of maintaining that code grows small.

Incomplete itemAllow pre/data/post files when schema and data are dumped separately, for performance reasons

Incomplete itemRefactor handling of database attributes between pg_dump and pg_dumpall

Currently only pg_dumpall emits database attributes, such as ALTER DATABASE SET commands and database-level GRANTs. Many people wish that pg_dump would do that. One proposal is to let pg_dump issue such commands if the -C switch was used, but it's unclear whether that will satisfy the demand.

Incomplete itemChange pg_dump so that a comment on the dumped database is applied to the loaded database, even if the database has a different name.

This will require new backend syntax, perhaps COMMENT ON CURRENT DATABASE. This is related to the previous item.

Incomplete itemAllow parallel restore of tar dumps

ecpg

Incomplete itemDocs

Document differences between ecpg and the SQL standard and information about the Informix-compatibility module.

Incomplete itemSolve cardinality > 1 for input descriptors / variables?

Incomplete itemAdd a semantic check level, e.g. check if a table really exists

Incomplete itemfix handling of DB attributes that are arrays

[D] Completed itemImplement SQLDA

add sqlda support to ecpg in both native and compatibility mode

Incomplete itemFix nested C comments

[E] Incomplete itemsqlwarn[6] should be 'W' if the PRECISION or SCALE value specified

Incomplete itemMake SET CONNECTION thread-aware, non-standard?

Incomplete itemAllow multidimensional arrays

Incomplete itemImplement COPY FROM STDIN

Incomplete itemProvide a way to specify size of a bytea parameter

[E] Incomplete itemFix small memory leaks in ecpg

Memory leaks in a short running application like ecpg are not really a problem, but make debugging more complicated

Incomplete itemAllow re-usage of cursor name variables

libpq

[D] Completed itemAdd PQescapeIdentifier()

Incomplete itemPrevent PQfnumber() from lowercasing unquoted column names

PQfnumber() should never have been doing lowercasing, but historically it has so we need a way to prevent it

Incomplete itemAllow statement results to be automatically batched to the client

Currently all statement results are transferred to the libpq client before libpq makes the results available to the application. This feature would allow the application to make use of the first result rows while the rest are transferred, or held on the server waiting for them to be requested by libpq. One complexity is that a statement like SELECT 1/col could error out mid-way through the result set.

Incomplete itemConsider disallowing multiple queries in PQexec() as an additional barrier to SQL injection attacks

Incomplete itemAdd PQexecf() that allows complex parameter substitution

Incomplete itemAdd SQLSTATE and severity to errors generated within libpq itself

Incomplete itemAdd code to detect client encoding and locale from the operating system environment

[D] Completed itemAdd keepalive support to libpq

Incomplete itemAdd support for interface/ipaddress binding to libpq

Triggers

Incomplete itemImprove storage of deferred trigger queue

Right now all deferred trigger information is stored in backend memory. This could exhaust memory for very large trigger queues. This item involves dumping large queues into files, or doing some kind of join to process all the triggers, some bulk operation, or a bitmap.

Incomplete itemAllow triggers to be disabled in only the current session.

This is currently possible by starting a multi-statement transaction, modifying the system tables, performing the desired SQL, restoring the system tables, and committing the transaction. ALTER TABLE ... TRIGGER requires a table lock so it is not ideal for this usage.

Incomplete itemWith disabled triggers, allow pg_dump to use ALTER TABLE ADD FOREIGN KEY

If the dump is known to be valid, allow foreign keys to be added without revalidating the data.

Incomplete itemAllow statement-level triggers to access modified rows

Incomplete itemWhen statement-level triggers are defined on a parent table, have them fire only on the parent table, and fire child table triggers only where appropriate

[D] Completed itemSupport triggers on columns

Incomplete itemAllow AFTER triggers on system tables

System tables are modified in many places in the backend without going through the executor and therefore not causing triggers to fire. To complete this item, the functions that modify system tables will have to fire triggers.

Incomplete itemTighten trigger permission checks

Incomplete itemAllow BEFORE INSERT triggers on views

Incomplete itemAdd database and transaction-level triggers

Incomplete itemReduce locking requirements for creating a trigger

Inheritance

Incomplete itemAllow inherited tables to inherit indexes, UNIQUE constraints, and primary/foreign keys

Incomplete itemHonor UNIQUE INDEX on base column in INSERTs/UPDATEs on inherited table, e.g. INSERT INTO inherit_table (unique_index_col) VALUES (dup) should fail

The main difficulty with this item is the problem of creating an index that can span multiple tables.

Incomplete itemDetermine whether ALTER TABLE / SET SCHEMA should work on inheritance hierarchies (and thus support ONLY). If yes, implement it.

Incomplete itemALTER TABLE variants sometimes support recursion and sometimes not, but this is poorly/not documented, and the ONLY marker would then be silently ignored. Clarify the documentation, and reject ONLY if it is not supported.

Indexes

Incomplete itemPrevent index uniqueness checks when UPDATE does not modify the column

Uniqueness (index) checks are done when updating a column even if the column is not modified by the UPDATE. However, HOT already short-circuits this in common cases, so more work might not be helpful.

Incomplete itemAllow the creation of on-disk bitmap indexes which can be quickly combined with other bitmap indexes

Such indexes could be more compact if there are only a few distinct values. Such indexes can also be compressed. Keeping such indexes updated can be costly.

Incomplete itemAllow accurate statistics to be collected on indexes with more than one column or expression indexes, perhaps using per-index statistics

Incomplete itemConsider having a larger statistics target for indexed columns and expression indexes.

Incomplete itemConsider smaller indexes that record a range of values per heap page, rather than having one index entry for every heap row

This is useful if the heap is clustered by the indexed values.

Incomplete itemAdd REINDEX CONCURRENTLY, like CREATE INDEX CONCURRENTLY

This is difficult because you must upgrade to an exclusive table lock to replace the existing index file. CREATE INDEX CONCURRENTLY does not have this complication. This would allow index compaction without downtime.

Incomplete itemAllow multiple indexes to be created concurrently, ideally via a single heap scan

pg_restore allows parallel index builds, but it is done via subprocesses, and there is no SQL interface for this.

Incomplete itemConsider sorting entries before inserting into btree index

Incomplete itemAllow index scans to return matching index keys, not just the matching heap locations

Incomplete itemAllow creation of an index that can do comparisons to test if a value is between two column values

Incomplete itemConsider using "effective_io_concurrency" for index scans

  • Currently only bitmap scans use this, which might be fine because most multi-row index scans use bitmap scans.

GIST

Incomplete itemAdd more GIST index support for geometric data types

Incomplete itemAllow GIST indexes to create certain complex index types, like digital trees (see Aoki)

Incomplete itemFix performance issues in contrib/seg and contrib/cube GiST support

GIN

Incomplete itemSupport empty indexed values (such as zero-element arrays) properly

Incomplete itemBehave correctly for cases where some elements of an indexed value are NULL

Incomplete itemSupport queries that require a full scan

Hash

Incomplete itemAdd UNIQUE capability to hash indexes

[D] Completed itemPack hash index buckets onto disk pages more efficiently

Currently only one hash bucket can be stored on a page. Ideally several hash buckets could be stored on a single page and greater granularity used for the hash algorithm.

However, the binary searching within a hash page probably renders this issue moot.

Incomplete itemAdd hash WAL logging for crash recovery

Incomplete itemAllow multi-column hash indexes

Catalogs

[D] Completed itemImprove performance of information_schema views

[D] Completed itemImprove information_schema's entries for precision and scale

Sorting

Incomplete itemConsider whether duplicate keys should be sorted by block/offset

Incomplete itemConsider being smarter about memory and external files used during sorts

Incomplete itemConsider detoasting keys before sorting

Fsync

Incomplete itemDetermine optimal fdatasync/fsync, O_SYNC/O_DSYNC options

Ideally this requires a separate test program that can be run at initdb time or optionally later. Consider O_SYNC when O_DIRECT exists.

Incomplete itemAdd program to test if fsync has a delay compared to non-fsync

Incomplete itemConsider sorting writes during checkpoint

Cache Usage

Incomplete itemSpeed up COUNT(*)

We could use a fixed row count and a +/- count to follow MVCC visibility rules, or a single cached value could be used and invalidated if anyone modifies the table. Another idea is to get a count directly from a unique index, but for this to be faster than a sequential scan it must avoid access to the heap to obtain tuple visibility information.

Incomplete itemProvide a way to calculate an "estimated COUNT(*)"

Perhaps by using the optimizer's cardinality estimates or random sampling.

Incomplete itemAllow data to be pulled directly from indexes

Currently indexes do not have enough tuple visibility information to allow data to be pulled from the index without also accessing the heap. One way to allow this is to set a bit on index tuples to indicate if a tuple is currently visible to all transactions when the first valid heap lookup happens. This bit would have to be cleared when a heap tuple is expired. Another idea is to maintain a bitmap of heap pages where all rows are visible to all backends, and allow index lookups to reference that bitmap to avoid heap lookups, perhaps the same bitmap we might add someday to determine which heap pages need vacuuming. Frequently accessed bitmaps would have to be stored in shared memory. One 8k page of bitmaps could track 512MB of heap pages. A third idea would be for a heap scan to check if all rows are visible and if so set a per-table flag which can be checked by index scans. Any change to the table would have to clear the flag. To detect changes during the heap scan a counter could be set at the start and checked at the end --- if it is the same, the table has not been modified --- any table change would increment the counter.

Incomplete itemConsider automatic caching of statements at various levels:

  • Parsed query tree
  • Query execute plan
  • Query results

Incomplete itemConsider increasing internal areas (NUM_CLOG_BUFFERS) when shared buffers is increased

Incomplete itemConsider decreasing the amount of memory used by PrivateRefCount

Incomplete itemConsider allowing higher priority queries to have referenced buffer cache pages stay in memory longer

Vacuum

[D] Completed itemImprove VACUUM FULL's speed when major data movement is needed

For large table adjustments during VACUUM FULL, it would be faster to cluster or reindex rather than update the indexes piecemeal as it does now. Also, this behavior tends to bloat the indexes.

Incomplete itemAuto-fill the free space map by scanning the buffer cache or by checking pages written by the background writer

Incomplete itemConsider having single-page pruning update the visibility map

Incomplete itemImprove tracking of total relation tuple counts now that vacuum doesn't always scan the whole heap

Incomplete itemBias FSM towards returning free space near the beginning of the heap file, in hopes that empty pages at the end can be truncated by VACUUM

Incomplete itemMake FSM return free space based on table clustering, to assist in maintaining clustering?

Incomplete itemConsider a more compact data representation for dead tuple locations within VACUUM

Incomplete itemProvide more information in order to improve user-side estimates of dead space bloat in relations

Auto-vacuum

[E] Incomplete itemIssue log message to suggest VACUUM FULL if a table is nearly empty?

Incomplete itemPrevent long-lived temporary tables from causing frozen-xid advancement starvation

The problem is that autovacuum cannot vacuum them to set frozen xids; only the session that created them can do that.

Incomplete itemPrevent autovacuum from running if an old transaction is still running from the last vacuum

Incomplete itemHave free space allocation bias away from using trailing table pages

This improves the chances of truncating the table during vacuum

Incomplete itemHave autoanalyze of parent tables occur when child tables are modified

Locking

Incomplete itemFix priority ordering of read and write light-weight locks

Incomplete itemFix problem when multiple subtransactions of the same outer transaction hold different types of locks, and one subtransaction aborts

Incomplete itemAllow UPDATEs on only non-referential integrity columns not to conflict with referential integrity locks

Incomplete itemAdd idle_in_transaction_timeout GUC so locks are not held for long periods of time

Incomplete itemImprove deadlock detection when a page cleaning lock conflicts with a shared buffer that is pinned

Incomplete itemDetect deadlocks involving LockBufferForCleanup()

Incomplete itemConsider a lock timeout parameter

Incomplete itemConsider improving serialized transaction behavior to avoid anomalies

Startup Time Improvements

Incomplete itemExperiment with multi-threaded backend for backend creation

This would prevent the overhead associated with process creation. Most operating systems have trivial process creation time compared to database startup overhead, but a few operating systems (Win32, Solaris) might benefit from threading. Also explore the idea of a single session using multiple threads to execute a statement faster.

Write-Ahead Log

Incomplete itemEliminate need to write full pages to WAL before page modification

Currently, to protect against partial disk page writes, we write full page images to WAL before they are modified so we can correct any partial page writes during recovery. These pages can also be eliminated from point-in-time archive files.

Incomplete itemWhen full page writes are off, write CRC to WAL and check file system blocks on recovery

If CRC check fails during recovery, remember the page in case a later CRC for that page properly matches.

Incomplete itemWrite full pages during file system write and not when the page is modified in the buffer cache

This allows most full page writes to happen in the background writer. It might cause problems for applying WAL on recovery into a partially-written page, but later the full page will be replaced from WAL.

[D] Completed itemAllow WAL traffic to be streamed to another server for stand-by replication

Incomplete itemReduce WAL traffic so only modified values are written rather than entire rows

Incomplete itemAllow WAL information to recover corrupted pg_controldata

Incomplete itemFind a way to reduce rotational delay when repeatedly writing last WAL page

Currently fsync of WAL requires the disk platter to perform a full rotation to fsync again. One idea is to write the WAL to different offsets that might reduce the rotational delay.

Incomplete itemAllow WAL logging to be turned off for a table, but the table might be dropped or truncated during crash recovery

Allow tables to bypass WAL writes and just fsync() dirty pages on commit. This should be implemented using ALTER TABLE, e.g. ALTER TABLE PERSISTENCE [ DROP | TRUNCATE | DEFAULT ]. Tables using non-default logging should not use referential integrity with default-logging tables. A table without dirty buffers during a crash could perhaps avoid the drop/truncate.

Incomplete itemAllow WAL logging to be turned off for a table, but the table would avoid being truncated/dropped

To do this, only a single writer can modify the table, and writes must happen only on new pages so the new pages can be removed during crash recovery. Readers can continue accessing the table. Such tables probably cannot have indexes. One complexity is the handling of indexes on TOAST tables.

Incomplete itemSpeed WAL recovery by allowing more than one page to be prefetched

This should be done utilizing the same infrastructure used for prefetching in general to avoid introducing complex error-prone code in WAL replay.

Incomplete itemImprove WAL concurrency by increasing lock granularity

Incomplete itemBe more aggressive about creating WAL files

Incomplete itemHave resource managers report the duration of their status changes

Incomplete itemMove pgfoundry's xlogdump to /contrib and have it rely more closely on the WAL backend code

Incomplete itemClose deleted WAL files held open in *nix by long-lived read-only backends

Optimizer / Executor

Incomplete itemImprove selectivity functions for geometric operators

Incomplete itemPrecompile SQL functions to avoid overhead

Incomplete itemCreate utility to compute accurate random_page_cost value

Incomplete itemConsider increasing the default values of from_collapse_limit, join_collapse_limit, and/or geqo_threshold

Incomplete itemImprove ability to display optimizer analysis using OPTIMIZER_DEBUG

Incomplete itemHave EXPLAIN ANALYZE issue NOTICE messages when the estimated and actual row counts differ by a specified percentage

Incomplete itemHave EXPLAIN ANALYZE report rows as floating-point numbers

Incomplete itemLog statements where the optimizer row estimates were dramatically different from the number of rows actually found?

Incomplete itemImprove how ANALYZE computes in-doubt tuples

Incomplete itemConsider compressed annealing to search for query plans

This might replace GEQO.

Incomplete itemConsider using a hash for joining to a large IN (VALUES ...) list

Incomplete itemAllow single batch hash joins to preserve outer pathkeys

Incomplete item"lazy" hash tables - look up only the tuples that are actually requested

Incomplete itemAvoid building the same hash table more than once during the same query

Incomplete itemAvoid hashing for distinct and then re-hashing for hash join

Incomplete itemAllow hashing to be used on arrays, if the element type is hashable

Incomplete itemImprove use of expression indexes for ORDER BY

Background Writer

Incomplete itemConsider having the background writer update the transaction status hint bits before writing out the page

Implementing this requires the background writer to have access to system catalogs and the transaction status log.

Incomplete itemConsider adding buffers the background writer finds reusable to the free list

Incomplete itemAutomatically tune bgwriter_delay based on activity rather then using a fixed interval

Incomplete itemConsider whether increasing BM_MAX_USAGE_COUNT improves performance

Incomplete itemTest to see if calling PreallocXlogFiles() from the background writer will help with WAL segment creation latency

Concurrent Use of Resources

Incomplete itemDo async I/O for faster random read-ahead of data

Async I/O allows multiple I/O requests to be sent to the disk with results coming back asynchronously.

The above patch is already applied as of 8.4, but it still remains to figure out how to handle plain indexscans effectively.

Incomplete itemExperiment with multi-threaded backend for better I/O utilization

This would allow a single query to make use of multiple I/O channels simultaneously. One idea is to create a background reader that can pre-fetch sequential and index scan pages needed by other backends. This could be expanded to allow concurrent reads from multiple devices in a partitioned table.

Incomplete itemExperiment with multi-threaded backend for better CPU utilization

This would allow several CPUs to be used for a single query, such as for sorting or query execution.

Incomplete itemSMP scalability improvements

TOAST

Incomplete itemAllow user configuration of TOAST thresholds

Incomplete itemReduce unnecessary cases of deTOASTing

Incomplete itemReduce costs of repeat de-TOASTing of values

Miscellaneous Performance

Incomplete itemUse mmap() rather than SYSV shared memory or to write WAL files?

This would remove the requirement for SYSV SHM but would introduce portability issues. Anonymous mmap (or mmap to /dev/zero) is required to prevent I/O overhead.

Incomplete itemConsider mmap()'ing files into a backend?

Doing I/O to large tables would consume a lot of address space or require frequent mapping/unmapping. Extending the file also causes mapping problems that might require mapping only individual pages, leading to thousands of mappings. Another problem is that there is no way to _prevent_ I/O to disk from the dirty shared buffers so changes could hit disk before WAL is written.

Incomplete itemAdd a script to ask system configuration questions and tune postgresql.conf

Incomplete itemConsider ways of storing rows more compactly on disk:

  • Reduce the row header size?
  • Consider reducing on-disk varlena length from four bytes to two because a heap row cannot be more than 64k in length

Incomplete itemConsider transaction start/end performance improvements

Incomplete itemAllow configuration of backend priorities via the operating system

Though backend priorities make priority inversion during lock waits possible, research shows that this is not a huge problem.

Incomplete itemConsider increasing the minimum allowed number of shared buffers

Incomplete itemConsider if CommandCounterIncrement() can avoid its AcceptInvalidationMessages() call

Incomplete itemConsider Cartesian joins when both relations are needed to form an indexscan qualification for a third relation

Incomplete itemConsider not storing a NULL bitmap on disk if all the NULLs are trailing

Incomplete itemSort large UPDATE/DELETEs so it is done in heap order

Incomplete itemAllow one transaction to see tuples using the snapshot of another transaction

This would assist multiple backends in working together.

Incomplete itemConsider decreasing the I/O caused by updating tuple hint bits

Incomplete itemAvoid the requirement of freezing pages that are infrequently modified

If all rows on a page are visible, it is possible to set a bit in the visibility map (once the visibility map is 100% reliable) and not need to freeze the page, avoiding a page rewrite

Incomplete itemAvoid reading in b-tree pages when replaying vacuum records in hot standby mode

Miscellaneous Other

Incomplete itemDeal with encoding issues for filenames in the server filesystem

Incomplete itemDeal with encoding issues in the output of localeconv()

Incomplete itemProvide schema name and other fields available from SQL GET DIAGNOSTICS in error reports

[E] Incomplete item Provide fallback_application_name in contrib/pgbench, oid2name, and dblink.

Source Code

Incomplete itemAdd use of 'const' for variables in source tree

[E] Incomplete itemRemove warnings created by -Wcast-align

Incomplete itemMove platform-specific ps status display info from ps_status.c to ports

Incomplete itemAdd optional CRC checksum to heap and index pages

One difficulty is how to prevent hint bit changes from affecting the computed CRC checksum.

Incomplete itemConsider a faster CRC32 algorithm

Incomplete itemAllow cross-compiling by generating the zic database on the target system

Incomplete itemImprove NLS maintenance of libpgport messages linked onto applications

Incomplete itemImprove the module installation experience (/contrib, etc)

Incomplete itemUse UTF8 encoding for NLS messages so all server encodings can read them properly

[D] Completed itemUpdate Bonjour to work with newer cross-platform SDK

Incomplete itemAllow creation of universal binaries for Darwin

Incomplete itemConsider GnuTLS if OpenSSL license becomes a problem

Incomplete itemConsider making NAMEDATALEN more configurable in future releases

Incomplete itemResearch use of signals and sleep wake ups

[D] Completed itemAdd automated check for invalid C++ source code constructs

Incomplete itemAllow C++ code to more easily access backend code

Incomplete itemConsider simplifying how memory context resets handle child contexts

Incomplete itemCreate three versions of libpgport to simplify client code

Incomplete itemImprove detection of shared memory segments being used by others by checking the SysV shared memory field 'nattch'

Incomplete itemImplement the non-threaded Avahi service discovery protocol

[D] Completed itemAllow table and index WITH options to be specified via hooks, for use with plugins like GiST index methods

Incomplete itemReduce data row alignment requirements on some 64-bit systems

Incomplete itemRestructure TOAST internal storage format for greater flexibility

[D] Completed itemAllow setting of system oids during object creation, for use by pg_migrator

Incomplete item Add regression tests for pg_dump/restore

Documentation

Incomplete itemConvert single quotes to apostrophes in the PDF documentation

Incomplete itemProvide a manpage for postgresql.conf

Incomplete itemChange the manpage-generating toolchain to use the new XML-based docbook2x tools

Incomplete itemConsider changing documentation format from SGML to XML

Incomplete itemDocument support for N' ' national character string literals, if it matches the SQL standard

/contrib/pg_upgrade

Incomplete itemRemove copy_dir() code, or use it

Incomplete itemHandle large object comments

This is difficult to do because the large object doesn't exist when --schema-only is loaded.

Incomplete itemConsider using pg_depend for checking object usage in version.c

Incomplete itemIf reindex is necessary, allow it to be done in parallel with pg_dump custom format

Incomplete itemMigrate pg_statistic by dumping it out as a flat file, so analyze is not necessary

pg_class.oid is not preserved so schema.tablename must be used.

Incomplete itemImprove testing, perhaps using the buildfarm

The buildfarm has access to multiple versions of PostgreSQL.

Windows

Incomplete itemRemove configure.in check for link failure when cause is found

Incomplete itemRemove readdir() errno patch when runtime/mingwex/dirent.c rev 1.4 is released

Incomplete itemAllow psql to use readline once non-US code pages work with backslashes

Incomplete itemFix problem with shared memory on the Win32 Terminal Server

[D] Completed itemDiagnose problem where shared memory can sometimes not be attached by postmaster children

Incomplete itemImprove signal handling

Incomplete itemConvert MSVC build system to remove most batch files

Incomplete itemSupport pgxs when using MSVC

Incomplete itemFix MSVC NLS support, like for to_char()

[D] Completed itemFix locale-aware handling (e.g. monetary) for specific server/client encoding combinations

Incomplete itemFind a correct rint() substitute on Windows

[D] Completed itemReduce compiler warnings on 64-bit Windows

Incomplete itemFix global namespace issues when using multiple terminal server sessions

Incomplete itemChange from the current autoconf/gmake build system to cmake

Incomplete itemImprove consistency of path separator usage

[D] Completed itemAllow compilation using MSVC 2008

Wire Protocol Changes

Incomplete itemAllow dynamic character set handling

Incomplete itemAdd decoded type, length, precision

Incomplete itemUse compression?

Incomplete itemUpdate clients to use data types, typmod, schema.table.column names of result sets using new statement protocol

Exotic Features

Incomplete itemAdd pre-parsing phase that converts non-ISO syntax to supported syntax

This could allow SQL written for other databases to run without modification.

Incomplete itemAllow plug-in modules to emulate features from other databases

Incomplete itemAdd features of Oracle-style packages

A package would be a schema with session-local variables, public/private functions, and initialization functions. It is also possible to implement these capabilities in any schema and not use a separate "packages" syntax at all.

Incomplete itemConsider allowing control of upper/lower case folding of unquoted identifiers

Incomplete itemAdd autonomous transactions

Incomplete itemGive query progress indication

Incomplete itemRethink our type system

Features We Do Not Want

Incomplete itemAll backends running as threads in a single process (not wanted)

This eliminates the process protection we get from the current setup. Thread creation is usually the same overhead as process creation on modern systems, so it seems unwise to use a pure threaded model, and MySQL and DB2 have demonstrated that threads introduce as many issues as they solve. Threading specific operations such as I/O, seq scans, and connection management has been discussed and will probably be implemented to enable specific performance features. Moving to a threaded engine would also require halting all other work on PostgreSQL for one to two years.

Incomplete itemOptimizer hints (not wanted)

Optimizer hints are used to work around problems in the optimizer and introduce upgrade and maintenance issues. We would rather have the problems reported and fixed. We have discussed a more sophisticated system of per-class cost adjustment instead, but a specification remains to be developed.

Incomplete itemEmbedded server (not wanted)

While PostgreSQL clients runs fine in limited-resource environments, the server requires multiple processes and a stable pool of resources to run reliably and efficiently. Stripping down the PostgreSQL server to run in the same process address space as the client application would add too much complexity and failure cases. Besides, there are several very mature embedded SQL databases already available.

Incomplete itemObfuscated function source code (not wanted)

Obfuscating function source code has minimal protective benefits because anyone with super-user access can find a way to view the code. At the same time, it would greatly complicate backups and other administrative tasks. To prevent non-super-users from viewing function source code, remove SELECT permission on pg_proc.

Incomplete itemIndeterminate behavior for the GROUP BY clause (not wanted)

At least one other database product allows specification of a subset of the result columns which GROUP BY would need to be able to provide predictable results; the server is free to return any value from the group. This is not viewed as a desirable feature.

Personal tools