5.10. Table Partitioning
PostgreSQL supports basic table partitioning. This section describes why and how to implement partitioning as part of your database design.
Partitioning refers to splitting what is logically one large table into smaller physical pieces. Partitioning can provide several benefits:
Query performance can be improved dramatically in certain situations, particularly when most of the heavily accessed rows of the table are in a single partition or a small number of partitions. The partitioning substitutes for leading columns of indexes, reducing index size and making it more likely that the heavily-used parts of the indexes fit in memory.
When queries or updates access a large percentage of a single partition, performance can be improved by taking advantage of sequential scan of that partition instead of using an index and random access reads scattered across the whole table.
Bulk loads and deletes can be accomplished by adding or removing partitions, if that requirement is planned into the partitioning design. Doing
ALTER TABLE DETACH PARTITIONor dropping an individual partition using
DROP TABLEis far faster than a bulk operation. These commands also entirely avoid the
VACUUMoverhead caused by a bulk
Seldom-used data can be migrated to cheaper and slower storage media.
The benefits will normally be worthwhile only when a table would otherwise be very large. The exact point at which a table will benefit from partitioning depends on the application, although a rule of thumb is that the size of the table should exceed the physical memory of the database server.
PostgreSQL offers built-in support for the following forms of partitioning:
- Range Partitioning
The table is partitioned into “ranges” defined by a key column or set of columns, with no overlap between the ranges of values assigned to different partitions. For example, one might partition by date ranges, or by ranges of identifiers for particular business objects.
- List Partitioning
The table is partitioned by explicitly listing which key values appear in each partition.
- Hash Partitioning
The table is partitioned by specifying a modulus and a remainder for each partition. Each partition will hold the rows for which the hash value of the partition key divided by the specified modulus will produce the specified remainder.
If your application needs to use other forms of partitioning not listed above, alternative methods such as inheritance and
UNION ALL views can be used instead. Such methods offer flexibility but do not have some of the performance benefits of built-in declarative partitioning.
5.10.2. Declarative Partitioning
PostgreSQL offers a way to specify how to divide a table into pieces called partitions. The table that is divided is referred to as a partitioned table. The specification consists of the partitioning method and a list of columns or expressions to be used as the partition key.
All rows inserted into a partitioned table will be routed to one of the partitions based on the value of the partition key. Each partition has a subset of the data defined by its partition bounds. The currently supported partitioning methods are range, list, and hash.
Partitions may themselves be defined as partitioned tables, using what is called sub-partitioning. Partitions may have their own indexes, constraints and default values, distinct from those of other partitions. See CREATE TABLE for more details on creating partitioned tables and partitions.
It is not possible to turn a regular table into a partitioned table or vice versa. However, it is possible to add a regular or partitioned table containing data as a partition of a partitioned table, or remove a partition from a partitioned table turning it into a standalone table; see ALTER TABLE to learn more about the
ATTACH PARTITION and
DETACH PARTITION sub-commands.
Individual partitions are linked to the partitioned table with inheritance behind-the-scenes; however, it is not possible to use some of the generic features of inheritance (discussed below) with declaratively partitioned tables or their partitions. For example, a partition cannot have any parents other than the partitioned table it is a partition of, nor can a regular table inherit from a partitioned table making the latter its parent. That means partitioned tables and their partitions do not participate in inheritance with regular tables. Since a partition hierarchy consisting of the partitioned table and its partitions is still an inheritance hierarchy, all the normal rules of inheritance apply as described in Section 5.9 with some exceptions, most notably:
NOT NULLconstraints of a partitioned table are always inherited by all its partitions.
CHECKconstraints that are marked
NO INHERITare not allowed to be created on partitioned tables.
ONLYto add or drop a constraint on only the partitioned table is supported as long as there are no partitions. Once partitions exist, using
ONLYwill result in an error as adding or dropping constraints on only the partitioned table, when partitions exist, is not supported. Instead, constraints on the partitions themselves can be added and (if they are not present in the parent table) dropped.
As a partitioned table does not have any data directly, attempts to use
ONLYon a partitioned table will always return an error.
Partitions cannot have columns that are not present in the parent. It is not possible to specify columns when creating partitions with
CREATE TABLE, nor is it possible to add columns to partitions after-the-fact using
ALTER TABLE. Tables may be added as a partition with
ALTER TABLE ... ATTACH PARTITIONonly if their columns exactly match the parent, including any
You cannot drop the
NOT NULLconstraint on a partition's column if the constraint is present in the parent table.
Partitions can also be foreign tables, although they have some limitations that normal tables do not; see CREATE FOREIGN TABLE for more information.
Updating the partition key of a row might cause it to be moved into a different partition where this row satisfies the partition bounds.
Suppose we are constructing a database for a large ice cream company. The company measures peak temperatures every day as well as ice cream sales in each region. Conceptually, we want a table like:
CREATE TABLE measurement ( city_id int not null, logdate date not null, peaktemp int, unitsales int );
We know that most queries will access just the last week's, month's or quarter's data, since the main use of this table will be to prepare online reports for management. To reduce the amount of old data that needs to be stored, we decide to only keep the most recent 3 years worth of data. At the beginning of each month we will remove the oldest month's data. In this situation we can use partitioning to help us meet all of our different requirements for the measurements table.
To use declarative partitioning in this case, use the following steps:
measurementtable as a partitioned table by specifying the
PARTITION BYclause, which includes the partitioning method (
RANGEin this case) and the list of column(s) to use as the partition key.
CREATE TABLE measurement ( city_id int not null, logdate date not null, peaktemp int, unitsales int ) PARTITION BY RANGE (logdate);
You may decide to use multiple columns in the partition key for range partitioning, if desired. Of course, this will often result in a larger number of partitions, each of which is individually smaller. On the other hand, using fewer columns may lead to a coarser-grained partitioning criteria with smaller number of partitions. A query accessing the partitioned table will have to scan fewer partitions if the conditions involve some or all of these columns. For example, consider a table range partitioned using columns
firstname(in that order) as the partition key.
Create partitions. Each partition's definition must specify the bounds that correspond to the partitioning method and partition key of the parent. Note that specifying bounds such that the new partition's values will overlap with those in one or more existing partitions will cause an error. Inserting data into the parent table that does not map to one of the existing partitions will cause an error; an appropriate partition must be added manually.
Partitions thus created are in every way normal PostgreSQL tables (or, possibly, foreign tables). It is possible to specify a tablespace and storage parameters for each partition separately.
It is not necessary to create table constraints describing partition boundary condition for partitions. Instead, partition constraints are generated implicitly from the partition bound specification whenever there is need to refer to them.
CREATE TABLE measurement_y2006m02 PARTITION OF measurement FOR VALUES FROM ('2006-02-01') TO ('2006-03-01'); CREATE TABLE measurement_y2006m03 PARTITION OF measurement FOR VALUES FROM ('2006-03-01') TO ('2006-04-01'); ... CREATE TABLE measurement_y2007m11 PARTITION OF measurement FOR VALUES FROM ('2007-11-01') TO ('2007-12-01'); CREATE TABLE measurement_y2007m12 PARTITION OF measurement FOR VALUES FROM ('2007-12-01') TO ('2008-01-01') TABLESPACE fasttablespace; CREATE TABLE measurement_y2008m01 PARTITION OF measurement FOR VALUES FROM ('2008-01-01') TO ('2008-02-01') WITH (parallel_workers = 4) TABLESPACE fasttablespace;
To implement sub-partitioning, specify the
PARTITION BYclause in the commands used to create individual partitions, for example:
CREATE TABLE measurement_y2006m02 PARTITION OF measurement FOR VALUES FROM ('2006-02-01') TO ('2006-03-01') PARTITION BY RANGE (peaktemp);
After creating partitions of
measurement_y2006m02, any data inserted into
measurementthat is mapped to
measurement_y2006m02(or data that is directly inserted into
measurement_y2006m02, provided it satisfies its partition constraint) will be further redirected to one of its partitions based on the
peaktempcolumn. The partition key specified may overlap with the parent's partition key, although care should be taken when specifying the bounds of a sub-partition such that the set of data it accepts constitutes a subset of what the partition's own bounds allows; the system does not try to check whether that's really the case.
Create an index on the key column(s), as well as any other indexes you might want, on the partitioned table. (The key index is not strictly necessary, but in most scenarios it is helpful.) This automatically creates one index on each partition, and any partitions you create or attach later will also contain the index.
CREATE INDEX ON measurement (logdate);
Ensure that the enable_partition_pruning configuration parameter is not disabled in
postgresql.conf. If it is, queries will not be optimized as desired.
In the above example we would be creating a new partition each month, so it might be wise to write a script that generates the required DDL automatically.
18.104.22.168. Partition Maintenance
Normally the set of partitions established when initially defining the table are not intended to remain static. It is common to want to remove old partitions of data and periodically add new partitions for new data. One of the most important advantages of partitioning is precisely that it allows this otherwise painful task to be executed nearly instantaneously by manipulating the partition structure, rather than physically moving large amounts of data around.
The simplest option for removing old data is to drop the partition that is no longer necessary:
DROP TABLE measurement_y2006m02;
This can very quickly delete millions of records because it doesn't have to individually delete every record. Note however that the above command requires taking an
ACCESS EXCLUSIVE lock on the parent table.
Another option that is often preferable is to remove the partition from the partitioned table but retain access to it as a table in its own right:
ALTER TABLE measurement DETACH PARTITION measurement_y2006m02;
This allows further operations to be performed on the data before it is dropped. For example, this is often a useful time to back up the data using
COPY, pg_dump, or similar tools. It might also be a useful time to aggregate data into smaller formats, perform other data manipulations, or run reports.
Similarly we can add a new partition to handle new data. We can create an empty partition in the partitioned table just as the original partitions were created above:
CREATE TABLE measurement_y2008m02 PARTITION OF measurement FOR VALUES FROM ('2008-02-01') TO ('2008-03-01') TABLESPACE fasttablespace;
As an alternative, it is sometimes more convenient to create the new table outside the partition structure, and make it a proper partition later. This allows the data to be loaded, checked, and transformed prior to it appearing in the partitioned table:
CREATE TABLE measurement_y2008m02 (LIKE measurement INCLUDING DEFAULTS INCLUDING CONSTRAINTS) TABLESPACE fasttablespace; ALTER TABLE measurement_y2008m02 ADD CONSTRAINT y2008m02 CHECK ( logdate >= DATE '2008-02-01' AND logdate < DATE '2008-03-01' ); \copy measurement_y2008m02 from 'measurement_y2008m02' -- possibly some other data preparation work ALTER TABLE measurement ATTACH PARTITION measurement_y2008m02 FOR VALUES FROM ('2008-02-01') TO ('2008-03-01' );
Before running the
ATTACH PARTITION command, it is recommended to create a
CHECK constraint on the table to be attached matching the desired partition constraint. That way, the system will be able to skip the scan to validate the implicit partition constraint. Without the
CHECK constraint, the table will be scanned to validate the partition constraint while holding an
ACCESS EXCLUSIVE lock on the parent table. It may be desired to drop the redundant
CHECK constraint after
ATTACH PARTITION is finished.
As explained above, it is possible to create indexes on partitioned tables and they are applied automatically to the entire hierarchy. This is very convenient, as not only the existing partitions will become indexed, but also any partitions that are created in the future will. One limitation is that it's not possible to use the
CONCURRENTLY qualifier when creating such a partitioned index. To overcome long lock times, it is possible to use
CREATE INDEX ON ONLY the partitioned table; such an index is marked invalid, and the partitions do not get the index applied automatically. The indexes on partitions can be created separately using
CONCURRENTLY, and later attached to the index on the parent using
ALTER INDEX .. ATTACH PARTITION. Once indexes for all partitions are attached to the parent index, the parent index is marked valid automatically. Example:
CREATE INDEX measurement_usls_idx ON ONLY measurement (unitsales); CREATE INDEX measurement_usls_200602_idx ON measurement_y2006m02 (unitsales); ALTER INDEX measurement_usls_idx ATTACH PARTITION measurement_usls_200602_idx; ...
This technique can be used with
PRIMARY KEY constraints too; the indexes are created implicitly when the constraint is created. Example:
ALTER TABLE ONLY measurement ADD UNIQUE (city_id, logdate); ALTER TABLE measurement_y2006m02 ADD UNIQUE (city_id, logdate); ALTER INDEX measurement_city_id_logdate_key ATTACH PARTITION measurement_y2006m02_city_id_logdate_key; ...
The following limitations apply to partitioned tables:
There is no way to create an exclusion constraint spanning all partitions; it is only possible to constrain each leaf partition individually.
Unique constraints on partitioned tables must include all the partition key columns. This limitation exists because PostgreSQL can only enforce uniqueness in each partition individually.
While primary keys are supported on partitioned tables, foreign keys referencing partitioned tables are not supported. (Foreign key references from a partitioned table to some other table are supported.)
BEFORE ROWtriggers, if necessary, must be defined on individual partitions, not the partitioned table.
Mixing temporary and permanent relations in the same partition tree is not allowed. Hence, if the partitioned table is permanent, so must be its partitions and likewise if the partitioned table is temporary. When using temporary relations, all members of the partition tree have to be from the same session.
5.10.3. Implementation Using Inheritance
While the built-in declarative partitioning is suitable for most common use cases, there are some circumstances where a more flexible approach may be useful. Partitioning can be implemented using table inheritance, which allows for several features not supported by declarative partitioning, such as:
For declarative partitioning, partitions must have exactly the same set of columns as the partitioned table, whereas with table inheritance, child tables may have extra columns not present in the parent.
Table inheritance allows for multiple inheritance.
Declarative partitioning only supports range, list and hash partitioning, whereas table inheritance allows data to be divided in a manner of the user's choosing. (Note, however, that if constraint exclusion is unable to prune child tables effectively, query performance might be poor.)
Some operations require a stronger lock when using declarative partitioning than when using table inheritance. For example, adding or removing a partition to or from a partitioned table requires taking an
ACCESS EXCLUSIVElock on the parent table, whereas a
SHARE UPDATE EXCLUSIVElock is enough in the case of regular inheritance.
We use the same
measurement table we used above. To implement partitioning using inheritance, use the following steps:
Create the “master” table, from which all of the “child” tables will inherit. This table will contain no data. Do not define any check constraints on this table, unless you intend them to be applied equally to all child tables. There is no point in defining any indexes or unique constraints on it, either. For our example, the master table is the
measurementtable as originally defined.
Create several “child” tables that each inherit from the master table. Normally, these tables will not add any columns to the set inherited from the master. Just as with declarative partitioning, these tables are in every way normal PostgreSQL tables (or foreign tables).
CREATE TABLE measurement_y2006m02 () INHERITS (measurement); CREATE TABLE measurement_y2006m03 () INHERITS (measurement); ... CREATE TABLE measurement_y2007m11 () INHERITS (measurement); CREATE TABLE measurement_y2007m12 () INHERITS (measurement); CREATE TABLE measurement_y2008m01 () INHERITS (measurement);
Add non-overlapping table constraints to the child tables to define the allowed key values in each.
Typical examples would be:
CHECK ( x = 1 ) CHECK ( county IN ( 'Oxfordshire', 'Buckinghamshire', 'Warwickshire' )) CHECK ( outletID >= 100 AND outletID < 200 )
Ensure that the constraints guarantee that there is no overlap between the key values permitted in different child tables. A common mistake is to set up range constraints like:
CHECK ( outletID BETWEEN 100 AND 200 ) CHECK ( outletID BETWEEN 200 AND 300 )
This is wrong since it is not clear which child table the key value 200 belongs in.
It would be better to instead create child tables as follows:
CREATE TABLE measurement_y2006m02 ( CHECK ( logdate >= DATE '2006-02-01' AND logdate < DATE '2006-03-01' ) ) INHERITS (measurement); CREATE TABLE measurement_y2006m03 ( CHECK ( logdate >= DATE '2006-03-01' AND logdate < DATE '2006-04-01' ) ) INHERITS (measurement); ... CREATE TABLE measurement_y2007m11 ( CHECK ( logdate >= DATE '2007-11-01' AND logdate < DATE '2007-12-01' ) ) INHERITS (measurement); CREATE TABLE measurement_y2007m12 ( CHECK ( logdate >= DATE '2007-12-01' AND logdate < DATE '2008-01-01' ) ) INHERITS (measurement); CREATE TABLE measurement_y2008m01 ( CHECK ( logdate >= DATE '2008-01-01' AND logdate < DATE '2008-02-01' ) ) INHERITS (measurement);
For each child table, create an index on the key column(s), as well as any other indexes you might want.
CREATE INDEX measurement_y2006m02_logdate ON measurement_y2006m02 (logdate); CREATE INDEX measurement_y2006m03_logdate ON measurement_y2006m03 (logdate); CREATE INDEX measurement_y2007m11_logdate ON measurement_y2007m11 (logdate); CREATE INDEX measurement_y2007m12_logdate ON measurement_y2007m12 (logdate); CREATE INDEX measurement_y2008m01_logdate ON measurement_y2008m01 (logdate);
We want our application to be able to say
INSERT INTO measurement ...and have the data be redirected into the appropriate child table. We can arrange that by attaching a suitable trigger function to the master table. If data will be added only to the latest child, we can use a very simple trigger function:
CREATE OR REPLACE FUNCTION measurement_insert_trigger() RETURNS TRIGGER AS $$ BEGIN INSERT INTO measurement_y2008m01 VALUES (NEW.*); RETURN NULL; END; $$ LANGUAGE plpgsql;
After creating the function, we create a trigger which calls the trigger function:
CREATE TRIGGER insert_measurement_trigger BEFORE INSERT ON measurement FOR EACH ROW EXECUTE FUNCTION measurement_insert_trigger();
We must redefine the trigger function each month so that it always points to the current child table. The trigger definition does not need to be updated, however.
We might want to insert data and have the server automatically locate the child table into which the row should be added. We could do this with a more complex trigger function, for example:
CREATE OR REPLACE FUNCTION measurement_insert_trigger() RETURNS TRIGGER AS $$ BEGIN IF ( NEW.logdate >= DATE '2006-02-01' AND NEW.logdate < DATE '2006-03-01' ) THEN INSERT INTO measurement_y2006m02 VALUES (NEW.*); ELSIF ( NEW.logdate >= DATE '2006-03-01' AND NEW.logdate < DATE '2006-04-01' ) THEN INSERT INTO measurement_y2006m03 VALUES (NEW.*); ... ELSIF ( NEW.logdate >= DATE '2008-01-01' AND NEW.logdate < DATE '2008-02-01' ) THEN INSERT INTO measurement_y2008m01 VALUES (NEW.*); ELSE RAISE EXCEPTION 'Date out of range. Fix the measurement_insert_trigger() function!'; END IF; RETURN NULL; END; $$ LANGUAGE plpgsql;
The trigger definition is the same as before. Note that each
IFtest must exactly match the
CHECKconstraint for its child table.
While this function is more complex than the single-month case, it doesn't need to be updated as often, since branches can be added in advance of being needed.
In practice, it might be best to check the newest child first, if most inserts go into that child. For simplicity, we have shown the trigger's tests in the same order as in other parts of this example.
A different approach to redirecting inserts into the appropriate child table is to set up rules, instead of a trigger, on the master table. For example:
CREATE RULE measurement_insert_y2006m02 AS ON INSERT TO measurement WHERE ( logdate >= DATE '2006-02-01' AND logdate < DATE '2006-03-01' ) DO INSTEAD INSERT INTO measurement_y2006m02 VALUES (NEW.*); ... CREATE RULE measurement_insert_y2008m01 AS ON INSERT TO measurement WHERE ( logdate >= DATE '2008-01-01' AND logdate < DATE '2008-02-01' ) DO INSTEAD INSERT INTO measurement_y2008m01 VALUES (NEW.*);
A rule has significantly more overhead than a trigger, but the overhead is paid once per query rather than once per row, so this method might be advantageous for bulk-insert situations. In most cases, however, the trigger method will offer better performance.
Be aware that
COPYignores rules. If you want to use
COPYto insert data, you'll need to copy into the correct child table rather than directly into the master.
COPYdoes fire triggers, so you can use it normally if you use the trigger approach.
Another disadvantage of the rule approach is that there is no simple way to force an error if the set of rules doesn't cover the insertion date; the data will silently go into the master table instead.
Ensure that the constraint_exclusion configuration parameter is not disabled in
postgresql.conf; otherwise child tables may be accessed unnecessarily.
As we can see, a complex table hierarchy could require a substantial amount of DDL. In the above example we would be creating a new child table each month, so it might be wise to write a script that generates the required DDL automatically.
22.214.171.124. Maintenance for Inheritance Partitioning
To remove old data quickly, simply drop the child table that is no longer necessary:
DROP TABLE measurement_y2006m02;
To remove the child table from the inheritance hierarchy table but retain access to it as a table in its own right:
ALTER TABLE measurement_y2006m02 NO INHERIT measurement;
To add a new child table to handle new data, create an empty child table just as the original children were created above:
CREATE TABLE measurement_y2008m02 ( CHECK ( logdate >= DATE '2008-02-01' AND logdate < DATE '2008-03-01' ) ) INHERITS (measurement);
Alternatively, one may want to create and populate the new child table before adding it to the table hierarchy. This could allow data to be loaded, checked, and transformed before being made visible to queries on the parent table.
CREATE TABLE measurement_y2008m02 (LIKE measurement INCLUDING DEFAULTS INCLUDING CONSTRAINTS); ALTER TABLE measurement_y2008m02 ADD CONSTRAINT y2008m02 CHECK ( logdate >= DATE '2008-02-01' AND logdate < DATE '2008-03-01' ); \copy measurement_y2008m02 from 'measurement_y2008m02' -- possibly some other data preparation work ALTER TABLE measurement_y2008m02 INHERIT measurement;
The following caveats apply to partitioning implemented using inheritance:
There is no automatic way to verify that all of the
CHECKconstraints are mutually exclusive. It is safer to create code that generates child tables and creates and/or modifies associated objects than to write each by hand.
The schemes shown here assume that the values of a row's key column(s) never change, or at least do not change enough to require it to move to another partition. An
UPDATEthat attempts to do that will fail because of the
CHECKconstraints. If you need to handle such cases, you can put suitable update triggers on the child tables, but it makes management of the structure much more complicated.
If you are using manual
ANALYZEcommands, don't forget that you need to run them on each child table individually. A command like:
will only process the master table.
ON CONFLICTclauses are unlikely to work as expected, as the
ON CONFLICTaction is only taken in case of unique violations on the specified target relation, not its child relations.
Triggers or rules will be needed to route rows to the desired child table, unless the application is explicitly aware of the partitioning scheme. Triggers may be complicated to write, and will be much slower than the tuple routing performed internally by declarative partitioning.
5.10.4. Partition Pruning
Partition pruning is a query optimization technique that improves performance for declaratively partitioned tables. As an example:
SET enable_partition_pruning = on; -- the default SELECT count(*) FROM measurement WHERE logdate >= DATE '2008-01-01';
Without partition pruning, the above query would scan each of the partitions of the
measurement table. With partition pruning enabled, the planner will examine the definition of each partition and prove that the partition need not be scanned because it could not contain any rows meeting the query's
WHERE clause. When the planner can prove this, it excludes (prunes) the partition from the query plan.
By using the EXPLAIN command and the enable_partition_pruning configuration parameter, it's possible to show the difference between a plan for which partitions have been pruned and one for which they have not. A typical unoptimized plan for this type of table setup is:
SET enable_partition_pruning = off; EXPLAIN SELECT count(*) FROM measurement WHERE logdate >= DATE '2008-01-01'; QUERY PLAN ----------------------------------------------------------------------------------- Aggregate (cost=188.76..188.77 rows=1 width=8) -> Append (cost=0.00..181.05 rows=3085 width=0) -> Seq Scan on measurement_y2006m02 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date) -> Seq Scan on measurement_y2006m03 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date) ... -> Seq Scan on measurement_y2007m11 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date) -> Seq Scan on measurement_y2007m12 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date) -> Seq Scan on measurement_y2008m01 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date)
Some or all of the partitions might use index scans instead of full-table sequential scans, but the point here is that there is no need to scan the older partitions at all to answer this query. When we enable partition pruning, we get a significantly cheaper plan that will deliver the same answer:
SET enable_partition_pruning = on; EXPLAIN SELECT count(*) FROM measurement WHERE logdate >= DATE '2008-01-01'; QUERY PLAN ----------------------------------------------------------------------------------- Aggregate (cost=37.75..37.76 rows=1 width=8) -> Append (cost=0.00..36.21 rows=617 width=0) -> Seq Scan on measurement_y2008m01 (cost=0.00..33.12 rows=617 width=0) Filter: (logdate >= '2008-01-01'::date)
Note that partition pruning is driven only by the constraints defined implicitly by the partition keys, not by the presence of indexes. Therefore it isn't necessary to define indexes on the key columns. Whether an index needs to be created for a given partition depends on whether you expect that queries that scan the partition will generally scan a large part of the partition or just a small part. An index will be helpful in the latter case but not the former.
Partition pruning can be performed not only during the planning of a given query, but also during its execution. This is useful as it can allow more partitions to be pruned when clauses contain expressions whose values are not known at query planning time; for example, parameters defined in a
PREPARE statement, using a value obtained from a subquery or using a parameterized value on the inner side of a nested loop join. Partition pruning during execution can be performed at any of the following times:
During initialization of the query plan. Partition pruning can be performed here for parameter values which are known during the initialization phase of execution. Partitions which are pruned during this stage will not show up in the query's
EXPLAIN ANALYZE. It is possible to determine the number of partitions which were removed during this phase by observing the “Subplans Removed” property in the
During actual execution of the query plan. Partition pruning may also be performed here to remove partitions using values which are only known during actual query execution. This includes values from subqueries and values from execution-time parameters such as those from parameterized nested loop joins. Since the value of these parameters may change many times during the execution of the query, partition pruning is performed whenever one of the execution parameters being used by partition pruning changes. Determining if partitions were pruned during this phase requires careful inspection of the
loopsproperty in the
EXPLAIN ANALYZEoutput. Subplans corresponding to different partitions may have different values for it depending on how many times each of them was pruned during execution. Some may be shown as
(never executed)if they were pruned every time.
Partition pruning can be disabled using the enable_partition_pruning setting.
Execution-time partition pruning currently only occurs for the
Append node type, not for
ModifyTable nodes. That is likely to be changed in a future release of PostgreSQL.
5.10.5. Partitioning and Constraint Exclusion
Constraint exclusion is a query optimization technique similar to partition pruning. While it is primarily used for partitioning implemented using the legacy inheritance method, it can be used for other purposes, including with declarative partitioning.
Constraint exclusion works in a very similar way to partition pruning, except that it uses each table's
CHECK constraints — which gives it its name — whereas partition pruning uses the table's partition bounds, which exist only in the case of declarative partitioning. Another difference is that constraint exclusion is only applied at plan time; there is no attempt to remove partitions at execution time.
The fact that constraint exclusion uses
CHECK constraints, which makes it slow compared to partition pruning, can sometimes be used as an advantage: because constraints can be defined even on declaratively-partitioned tables, in addition to their internal partition bounds, constraint exclusion may be able to elide additional partitions from the query plan.
The default (and recommended) setting of constraint_exclusion is neither
off, but an intermediate setting called
partition, which causes the technique to be applied only to queries that are likely to be working on inheritance partitioned tables. The
on setting causes the planner to examine
CHECK constraints in all queries, even simple ones that are unlikely to benefit.
The following caveats apply to constraint exclusion:
Constraint exclusion is only applied during query planning, unlike partition pruning, which can also be applied during query execution.
Constraint exclusion only works when the query's
WHEREclause contains constants (or externally supplied parameters). For example, a comparison against a non-immutable function such as
CURRENT_TIMESTAMPcannot be optimized, since the planner cannot know which child table the function's value might fall into at run time.
Keep the partitioning constraints simple, else the planner may not be able to prove that child tables might not need to be visited. Use simple equality conditions for list partitioning, or simple range tests for range partitioning, as illustrated in the preceding examples. A good rule of thumb is that partitioning constraints should contain only comparisons of the partitioning column(s) to constants using B-tree-indexable operators, because only B-tree-indexable column(s) are allowed in the partition key.
All constraints on all children of the parent table are examined during constraint exclusion, so large numbers of children are likely to increase query planning time considerably. So the legacy inheritance based partitioning will work well with up to perhaps a hundred child tables; don't try to use many thousands of children.
5.10.6. Declarative Partitioning Best Practices
The choice of how to partition a table should be made carefully as the performance of query planning and execution can be negatively affected by poor design.
One of the most critical design decisions will be the column or columns by which you partition your data. Often the best choice will be to partition by the column or set of columns which most commonly appear in
WHERE clauses of queries being executed on the partitioned table.
WHERE clause items that match and are compatible with the partition key can be used to prune unneeded partitions. However, you may be forced into making other decisions by requirements for the
PRIMARY KEY or a
UNIQUE constraint. Removal of unwanted data is also a factor to consider when planning your partitioning strategy. An entire partition can be detached fairly quickly, so it may be beneficial to design the partition strategy in such a way that all data to be removed at once is located in a single partition.
Choosing the target number of partitions that the table should be divided into is also a critical decision to make. Not having enough partitions may mean that indexes remain too large and that data locality remains poor which could result in low cache hit ratios. However, dividing the table into too many partitions can also cause issues. Too many partitions can mean longer query planning times and higher memory consumption during both query planning and execution. When choosing how to partition your table, it's also important to consider what changes may occur in the future. For example, if you choose to have one partition per customer and you currently have a small number of large customers, consider the implications if in several years you instead find yourself with a large number of small customers. In this case, it may be better to choose to partition by
HASH and choose a reasonable number of partitions rather than trying to partition by
LIST and hoping that the number of customers does not increase beyond what it is practical to partition the data by.
Sub-partitioning can be useful to further divide partitions that are expected to become larger than other partitions, although excessive sub-partitioning can easily lead to large numbers of partitions and can cause the same problems mentioned in the preceding paragraph.
It is also important to consider the overhead of partitioning during query planning and execution. The query planner is generally able to handle partition hierarchies with up to a few hundred partitions fairly well, provided that typical queries allow the query planner to prune all but a small number of partitions. Planning times become longer and memory consumption becomes higher as more partitions are added. This is particularly true for the
DELETE commands. Another reason to be concerned about having a large number of partitions is that the server's memory consumption may grow significantly over a period of time, especially if many sessions touch large numbers of partitions. That's because each partition requires its metadata to be loaded into the local memory of each session that touches it.
With data warehouse type workloads, it can make sense to use a larger number of partitions than with an OLTP type workload. Generally, in data warehouses, query planning time is less of a concern as the majority of processing time is spent during query execution. With either of these two types of workload, it is important to make the right decisions early, as re-partitioning large quantities of data can be painfully slow. Simulations of the intended workload are often beneficial for optimizing the partitioning strategy. Never assume that more partitions are better than fewer partitions and vice-versa.