5.11. Monitoring Database Activity

A database administrator frequently wonders, What is the system doing right now? This chapter discusses how to find that out.

Several tools are available for monitoring database activity and analyzing performance. Most of this chapter is devoted to describing PostgreSQL’s cumulative statistics system, but one should not neglect regular Unix monitoring programs such as ps, top, iostat, and vmstat. Also, once one has identified a poorly-performing query, further investigation might be needed using PostgreSQL’s linkend=»sql-explain»>**EXPLAIN** command. using-explain discusses EXPLAIN and other methods for understanding the behavior of an individual query.

5.11.1. Standard Unix Tools

On most Unix platforms, PostgreSQL modifies its command title as reported by ps, so that individual server processes can readily be identified. A sample display is

$ ps auxww | grep ^postgres
postgres  15551  0.0  0.1  57536  7132 pts/0    S    18:02   0:00 postgres -i
postgres  15554  0.0  0.0  57536  1184 ?        Ss   18:02   0:00 postgres: background writer
postgres  15555  0.0  0.0  57536   916 ?        Ss   18:02   0:00 postgres: checkpointer
postgres  15556  0.0  0.0  57536   916 ?        Ss   18:02   0:00 postgres: walwriter
postgres  15557  0.0  0.0  58504  2244 ?        Ss   18:02   0:00 postgres: autovacuum launcher
postgres  15582  0.0  0.0  58772  3080 ?        Ss   18:04   0:00 postgres: joe runbug 127.0.0.1 idle
postgres  15606  0.0  0.0  58772  3052 ?        Ss   18:07   0:00 postgres: tgl regression [local] SELECT waiting
postgres  15610  0.0  0.0  58772  3056 ?        Ss   18:07   0:00 postgres: tgl regression [local] idle in transaction

(The appropriate invocation of ps varies across different platforms, as do the details of what is shown. This example is from a recent Linux system.) The first process listed here is the primary server process. The command arguments shown for it are the same ones used when it was launched. The next four processes are background worker processes automatically launched by the primary process. (The autovacuum launcher process will not be present if you have set the system not to run autovacuum.) Each of the remaining processes is a server process handling one client connection. Each such process sets its command line display in the form

postgres: user database host activity

The user, database, and (client) host items remain the same for the life of the client connection, but the activity indicator changes. The activity can be idle (i.e., waiting for a client command), idle in transaction (waiting for client inside a BEGIN block), or a command type name such as SELECT. Also, waiting is appended if the server process is presently waiting on a lock held by another session. In the above example we can infer that process 15606 is waiting for process 15610 to complete its transaction and thereby release some lock. (Process 15610 must be the blocker, because there is no other active session. In more complicated cases it would be necessary to look into the

linkend=»view-pg-locks»>**pg_locks**

system view to determine who is blocking whom.)

If guc-cluster-name has been configured the cluster name will also be shown in ps output:

$ psql -c 'SHOW cluster_name'
 cluster_name
--------------
 server1
(1 row)

$ ps aux|grep server1
postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: server1: background writer
...

If you have turned off guc-update-process-title then the activity indicator is not updated; the process title is set only once when a new process is launched. On some platforms this saves a measurable amount of per-command overhead; on others it’s insignificant.

Совет

Solaris requires special handling. You must use /usr/ucb/ps, rather than /bin/ps. You also must use two w flags, not just one. In addition, your original invocation of the postgres command must have a shorter ps status display than that provided by each server process. If you fail to do all three things, the ps output for each server process will be the original postgres command line.

5.11.2. The Cumulative Statistics System

PostgreSQL’s cumulative statistics system supports collection and reporting of information about server activity. Presently, accesses to tables and indexes in both disk-block and individual-row terms are counted. The total number of rows in each table, and information about vacuum and analyze actions for each table are also counted. If enabled, calls to user-defined functions and the total time spent in each one are counted as well.

PostgreSQL also supports reporting dynamic information about exactly what is going on in the system right now, such as the exact command currently being executed by other server processes, and which other connections exist in the system. This facility is independent of the cumulative statistics system.

5.11.2.1. Statistics Collection Configuration

Since collection of statistics adds some overhead to query execution, the system can be configured to collect or not collect information. This is controlled by configuration parameters that are normally set in postgresql.conf. (See runtime-config for details about setting configuration parameters.)

The parameter guc-track-activities enables monitoring of the current command being executed by any server process.

The parameter guc-track-counts controls whether cumulative statistics are collected about table and index accesses.

The parameter guc-track-functions enables tracking of usage of user-defined functions.

The parameter guc-track-io-timing enables monitoring of block read and write times.

The parameter guc-track-wal-io-timing enables monitoring of WAL write times.

Normally these parameters are set in postgresql.conf so that they apply to all server processes, but it is possible to turn them on or off in individual sessions using the sql-set command. (To prevent ordinary users from hiding their activity from the administrator, only superusers are allowed to change these parameters with SET.)

Cumulative statistics are collected in shared memory. Every PostgreSQL process collects statistics locally, then updates the shared data at appropriate intervals. When a server, including a physical replica, shuts down cleanly, a permanent copy of the statistics data is stored in the pg_stat subdirectory, so that statistics can be retained across server restarts. In contrast, when starting from an unclean shutdown (e.g., after an immediate shutdown, a server crash, starting from a base backup, and point-in-time recovery), all statistics counters are reset.

5.11.2.2. Viewing Statistics

Several predefined views, listed in monitoring-stats-dynamic-views-table, are available to show the current state of the system. There are also several other views, listed in monitoring-stats-views-table, available to show the accumulated statistics. Alternatively, one can build custom views using the underlying cumulative statistics functions, as discussed in monitoring-stats-functions.

When using the cumulative statistics views and functions to monitor collected data, it is important to realize that the information does not update instantaneously. Each individual server process flushes out accumulated statistics to shared memory just before going idle, but not more frequently than once per PGSTAT_MIN_INTERVAL milliseconds (1 second unless altered while building the server); so a query or transaction still in progress does not affect the displayed totals and the displayed information lags behind actual activity. However, current-query information collected by track_activities is always up-to-date.

Another important point is that when a server process is asked to display any of the accumulated statistics, accessed values are cached until the end of its current transaction in the default configuration. So the statistics will show static information as long as you continue the current transaction. Similarly, information about the current queries of all sessions is collected when any such information is first requested within a transaction, and the same information will be displayed throughout the transaction. This is a feature, not a bug, because it allows you to perform several queries on the statistics and correlate the results without worrying that the numbers are changing underneath you.

When analyzing statistics interactively, or with expensive queries, the time delta between accesses to individual statistics can lead to significant skew in the cached statistics. To minimize skew, stats_fetch_consistency can be set to snapshot, at the price of increased memory usage for caching not-needed statistics data. Conversely, if it’s known that statistics are only accessed once, caching accessed statistics is unnecessary and can be avoided by setting stats_fetch_consistency to none.

You can invoke **pg_stat_clear_snapshot**() to discard the current transaction’s statistics snapshot or cached values (if any). The next use of statistical information will (when in snapshot mode) cause a new snapshot to be built or (when in cache mode) accessed statistics to be cached.

A transaction can also see its own statistics (not yet flushed out to the shared memory statistics) in the views pg_stat_xact_all_tables, pg_stat_xact_sys_tables, pg_stat_xact_user_tables, and pg_stat_xact_user_functions. These numbers do not act as stated above; instead they update continuously throughout the transaction.

Some of the information in the dynamic statistics views shown in monitoring-stats-dynamic-views-table is security restricted. Ordinary users can only see all the information about their own sessions (sessions belonging to a role that they are a member of). In rows about other sessions, many columns will be null. Note, however, that the existence of a session and its general properties such as its sessions user and database are visible to all users. Superusers and roles with privileges of built-in role pg_read_all_stats (see also predefined-roles) can see all the information about all sessions.

Dynamic Statistics Views

Collected Statistics Views

The per-index statistics are particularly useful to determine which indexes are being used and how effective they are.

The pg_statio_ views are primarily useful to determine the effectiveness of the buffer cache. When the number of actual disk reads is much smaller than the number of buffer hits, then the cache is satisfying most read requests without invoking a kernel call. However, these statistics do not give the entire story: due to the way in which PostgreSQL handles disk I/O, data that is not in the PostgreSQL buffer cache might still reside in the kernel’s I/O cache, and might therefore still be fetched without requiring a physical read. Users interested in obtaining more detailed information on PostgreSQL I/O behavior are advised to use the PostgreSQL statistics views in combination with operating system utilities that allow insight into the kernel’s handling of I/O.

5.11.2.3. pg_stat_activity

The pg_stat_activity view will have one row per server process, showing information related to the current activity of that process.

pg_stat_activity View

Column Type Description

datid oid OID of the database this backend is connected to

datname name Name of the database this backend is connected to

pid integer Process ID of this backend

leader_pid integer Process ID of the parallel group leader, if this process is a parallel query worker. NULL if this process is a parallel group leader or does not participate in parallel query.

usesysid oid OID of the user logged into this backend

usename name Name of the user logged into this backend

application_name text Name of the application that is connected to this backend

client_addr inet IP address of the client connected to this backend. If this field is null, it indicates either that the client is connected via a Unix socket on the server machine or that this is an internal process such as autovacuum.

client_hostname text Host name of the connected client, as reported by a reverse DNS lookup of client_addr. This field will only be non-null for IP connections, and only when guc-log-hostname is enabled.

client_port integer TCP port number that the client is using for communication with this backend, or -1 if a Unix socket is used. If this field is null, it indicates that this is an internal server process.

backend_start timestamp with time zone Time when this process was started. For client backends, this is the time the client connected to the server.

xact_start timestamp with time zone Time when this process“ current transaction was started, or null if no transaction is active. If the current query is the first of its transaction, this column is equal to the query_start column.

query_start timestamp with time zone Time when the currently active query was started, or if state is not active, when the last query was started

state_change timestamp with time zone Time when the state was last changed

wait_event_type text The type of event for which the backend is waiting, if any; otherwise NULL. See wait-event-table.

wait_event text Wait event name if backend is currently waiting, otherwise NULL. See wait-event-activity- table through wait-event-timeout-t able.

state text Current overall state of this backend. Possible values are: #. active: The backend is executing a query. #. idle: The backend is waiting for a new client command. #. idle in transaction: The backend is in a transaction, but is not currently executing a query. #. idle in transaction (aborted): This state is similar to idle in transaction, except one of the statements in the transaction caused an error. #. fastpath function call: The backend is executing a fast-path function. #. disabled: This state is reported if guc-track-activities is disabled in this backend.

backend_xid xid Top-level transaction identifier of this backend, if any.

backend_xmin xid The current backend’s xmin horizon.

query_id bigint Identifier of this backend’s most recent query. If state is active this field shows the identifier of the currently executing query. In all other states, it shows the identifier of last query that was executed. Query identifiers are not computed by default so this field will be null unless guc-compute-query-id parameter is enabled or a third-party module that computes query identifiers is configured.

query text Text of this backend’s most recent query. If state is active this field shows the currently executing query. In all other states, it shows the last query that was executed. By default the query text is truncated at 1024 bytes; this value can be changed via the parameter guc-track-activity-q uery-size.

backend_type text Type of current backend. Possible types are autovacuum launcher, autovacuum worker, logical replication launcher, logical replication worker, parallel worker, background writer, client backend, checkpointer, archiver, startup, walreceiver, walsender and walwriter. In addition, background workers registered by extensions may have additional types.

Примечание

The wait_event and state columns are independent. If a backend is in the active state, it may or may not be waiting on some event. If the state is active and wait_event is non-null, it means that a query is being executed, but is being blocked somewhere in the system.

Wait Event Types

Wait Events of Type Activity

Wait Events of Type BufferPin

Wait Events of Type Client

Wait Events of Type Extension

Wait Events of Type IO

Wait Events of Type IPC

Wait Events of Type Lock

Wait Events of Type LWLock

Примечание

Extensions can add LWLock types to the list shown in wait-event-lwlock-table. In some cases, the name assigned by an extension will not be available in all server processes; so an LWLock wait event might be reported as just extension rather than the extension-assigned name.

Wait Events of Type Timeout

Here is an example of how wait events can be viewed:

SELECT pid, wait_event_type, wait_event FROM pg_stat_activity WHERE wait_event is NOT NULL;
 pid  | wait_event_type | wait_event
------+-----------------+------------
 2540 | Lock            | relation
 6644 | LWLock          | ProcArray
(2 rows)

5.11.2.4. pg_stat_replication

The pg_stat_replication view will contain one row per WAL sender process, showing statistics about replication to that sender’s connected standby server. Only directly connected standbys are listed; no information is available about downstream standby servers.

pg_stat_replication View

Column Type Description

pid integer Process ID of a WAL sender process

usesysid oid OID of the user logged into this WAL sender process

usename name Name of the user logged into this WAL sender process

application_name text Name of the application that is connected to this WAL sender

client_addr inet IP address of the client connected to this WAL sender. If this field is null, it indicates that the client is connected via a Unix socket on the server machine.

client_hostname text Host name of the connected client, as reported by a reverse DNS lookup of client_addr. This field will only be non-null for IP connections, and only when guc-log-hostname is enabled.

client_port integer TCP port number that the client is using for communication with this WAL sender, or -1 if a Unix socket is used

backend_start timestamp with time zone Time when this process was started, i.e., when the client connected to this WAL sender

backend_xmin xid This standby’s xmin horizon reported by guc-hot-standby-feed back.

state text Current WAL sender state. Possible values are: #. startup: This WAL sender is starting up. #. catchup: This WAL sender’s connected standby is catching up with the primary. #. streaming: This WAL sender is streaming changes after its connected standby server has caught up with the primary. #. backup: This WAL sender is sending a backup. #. stopping: This WAL sender is stopping.

sent_lsn pg_lsn Last write-ahead log location sent on this connection

write_lsn pg_lsn Last write-ahead log location written to disk by this standby server

flush_lsn pg_lsn Last write-ahead log location flushed to disk by this standby server

replay_lsn pg_lsn Last write-ahead log location replayed into the database on this standby server

write_lag interval Time elapsed between flushing recent WAL locally and receiving notification that this standby server has written it (but not yet flushed it or applied it). This can be used to gauge the delay that synchronous_commit ** level **remote_write incurred while committing if this server was configured as a synchronous standby.

flush_lag interval Time elapsed between flushing recent WAL locally and receiving notification that this standby server has written and flushed it (but not yet applied it). This can be used to gauge the delay that synchronous_commit ** level **on incurred while committing if this server was configured as a synchronous standby.

replay_lag interval Time elapsed between flushing recent WAL locally and receiving notification that this standby server has written, flushed and applied it. This can be used to gauge the delay that synchronous_commit ** level **remote_apply incurred while committing if this server was configured as a synchronous standby.

sync_priority integer Priority of this standby server for being chosen as the synchronous standby in a priority-based synchronous replication. This has no effect in a quorum-based synchronous replication.

sync_state text Synchronous state of this standby server. Possible values are: #. async: This standby server is asynchronous. #. potential: This standby server is now asynchronous, but can potentially become synchronous if one of current synchronous ones fails. #. sync: This standby server is synchronous. #. quorum: This standby server is considered as a candidate for quorum standbys.

reply_time timestamp with time zone Send time of last reply message received from standby server

The lag times reported in the pg_stat_replication view are measurements of the time taken for recent WAL to be written, flushed and replayed and for the sender to know about it. These times represent the commit delay that was (or would have been) introduced by each synchronous commit level, if the remote server was configured as a synchronous standby. For an asynchronous standby, the replay_lag column approximates the delay before recent transactions became visible to queries. If the standby server has entirely caught up with the sending server and there is no more WAL activity, the most recently measured lag times will continue to be displayed for a short time and then show NULL.

Lag times work automatically for physical replication. Logical decoding plugins may optionally emit tracking messages; if they do not, the tracking mechanism will simply display NULL lag.

Примечание

The reported lag times are not predictions of how long it will take for the standby to catch up with the sending server assuming the current rate of replay. Such a system would show similar times while new WAL is being generated, but would differ when the sender becomes idle. In particular, when the standby has caught up completely, pg_stat_replication shows the time taken to write, flush and replay the most recent reported WAL location rather than zero as some users might expect. This is consistent with the goal of measuring synchronous commit and transaction visibility delays for recent write transactions. To reduce confusion for users expecting a different model of lag, the lag columns revert to NULL after a short time on a fully replayed idle system. Monitoring systems should choose whether to represent this as missing data, zero or continue to display the last known value.

5.11.2.5. pg_stat_replication_slots

The pg_stat_replication_slots view will contain one row per logical replication slot, showing statistics about its usage.

pg_stat_replication_slots View

Column Type Description

slot_name text A unique, cluster-wide identifier for the replication slot

spill_txns bigint Number of transactions spilled to disk once the memory used by logical decoding to decode changes from WAL has exceeded logical_decoding_w ork_mem. The counter gets incremented for both top-level transactions and subtransactions.

spill_count bigint Number of times transactions were spilled to disk while decoding changes from WAL for this slot. This counter is incremented each time a transaction is spilled, and the same transaction may be spilled multiple times.

spill_bytes bigint Amount of decoded transaction data spilled to disk while performing decoding of changes from WAL for this slot. This and other spill counters can be used to gauge the I/O which occurred during logical decoding and allow tuning logical_decoding_w ork_mem.

stream_txns bigint Number of in-progress transactions streamed to the decoding output plugin after the memory used by logical decoding to decode changes from WAL for this slot has exceeded logical_decoding_w ork_mem. Streaming only works with top-level transactions (subtransactions can’t be streamed independently), so the counter is not incremented for subtransactions.

stream_count****bi gint Number of times in-progress transactions were streamed to the decoding output plugin while decoding changes from WAL for this slot. This counter is incremented each time a transaction is streamed, and the same transaction may be streamed multiple times.

stream_bytes****bi gint Amount of transaction data decoded for streaming in-progress transactions to the decoding output plugin while decoding changes from WAL for this slot. This and other streaming counters for this slot can be used to tune logical_decoding_w ork_mem.

total_txns bigint Number of decoded transactions sent to the decoding output plugin for this slot. This counts top-level transactions only, and is not incremented for subtransactions. Note that this includes the transactions that are streamed and/or spilled.

total_bytes****big int Amount of transaction data decoded for sending transactions to the decoding output plugin while decoding changes from WAL for this slot. Note that this includes data that is streamed and/or spilled.

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.6. pg_stat_wal_receiver

The pg_stat_wal_receiver view will contain only one row, showing statistics about the WAL receiver from that receiver’s connected server.

pg_stat_wal_receiver View

Column Type Description

pid integer Process ID of the WAL receiver process

status text Activity status of the WAL receiver process

receive_start_lsn* * **pg_lsn First write-ahead log location used when WAL receiver is started

receive_start_tli* * **integer First timeline number used when WAL receiver is started

written_lsn pg_lsn Last write-ahead log location already received and written to disk, but not flushed. This should not be used for data integrity checks.

flushed_lsn pg_lsn Last write-ahead log location already received and flushed to disk, the initial value of this field being the first log location used when WAL receiver is started

received_tli integer Timeline number of last write-ahead log location received and flushed to disk, the initial value of this field being the timeline number of the first log location used when WAL receiver is started

last_msg_send_time ** **timestamp with time zone Send time of last message received from origin WAL sender

last_msg_receipt_t ime timestamp with time zone Receipt time of last message received from origin WAL sender

latest_end_lsn pg_lsn Last write-ahead log location reported to origin WAL sender

latest_end_time timestamp with time zone Time of last write-ahead log location reported to origin WAL sender

slot_name text Replication slot name used by this WAL receiver

sender_host text Host of the PostgreSQL instance this WAL receiver is connected to. This can be a host name, an IP address, or a directory path if the connection is via Unix socket. (The path case can be distinguished because it will always be an absolute path, beginning with /.)

sender_port integer Port number of the PostgreSQL instance this WAL receiver is connected to.

conninfo text Connection string used by this WAL receiver, with security-sensitive fields obfuscated.

5.11.2.7. pg_stat_recovery_prefetch

The pg_stat_recovery_prefetch view will contain only one row. The columns wal_distance, block_distance and io_depth show current values, and the other columns show cumulative counters that can be reset with the pg_stat_reset_shared function.

pg_stat_recovery_prefetch View

Column Type Description

stats_reset timestamp with time zone Time at which these statistics were last reset

prefetch bigint Number of blocks prefetched because they were not in the buffer pool

hit bigint Number of blocks not prefetched because they were already in the buffer pool

skip_init bigint Number of blocks not prefetched because they would be zero-initialized

skip_new bigint Number of blocks not prefetched because they didn’t exist yet

skip_fpw bigint Number of blocks not prefetched because a full page image was included in the WAL

skip_rep bigint Number of blocks not prefetched because they were already recently prefetched

wal_distance int How many bytes ahead the prefetcher is looking

block_distance int How many blocks ahead the prefetcher is looking

io_depth int How many prefetches have been initiated but are not yet known to have completed

5.11.2.8. pg_stat_subscription

pg_stat_subscription View

Column Type Description

subid oid OID of the subscription

subname name Name of the subscription

pid integer Process ID of the subscription worker process

relid oid OID of the relation that the worker is synchronizing; null for the main apply worker

received_lsn pg_lsn Last write-ahead log location received, the initial value of this field being 0

last_msg_send_time ** **timestamp with time zone Send time of last message received from origin WAL sender

last_msg_receipt_t ime timestamp with time zone Receipt time of last message received from origin WAL sender

latest_end_lsn pg_lsn Last write-ahead log location reported to origin WAL sender

latest_end_time timestamp with time zone Time of last write-ahead log location reported to origin WAL sender

5.11.2.9. pg_stat_subscription_stats

The pg_stat_subscription_stats view will contain one row per subscription.

pg_stat_subscription_stats View

Column Type Description

subid oid OID of the subscription

subname name Name of the subscription

apply_error_count* * **bigint Number of times an error occurred while applying changes

sync_error_count bigint Number of times an error occurred during the initial table synchronization

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.10. pg_stat_ssl

The pg_stat_ssl view will contain one row per backend or WAL sender process, showing statistics about SSL usage on this connection. It can be joined to pg_stat_activity or pg_stat_replication on the pid column to get more details about the connection.

pg_stat_ssl View

Column Type Description

pid integer Process ID of a backend or WAL sender process

ssl boolean True if SSL is used on this connection

version text Version of SSL in use, or NULL if SSL is not in use on this connection

cipher text Name of SSL cipher in use, or NULL if SSL is not in use on this connection

bits integer Number of bits in the encryption algorithm used, or NULL if SSL is not used on this connection

client_dn text Distinguished Name (DN) field from the client certificate used, or NULL if no client certificate was supplied or if SSL is not in use on this connection. This field is truncated if the DN field is longer than NAMEDATALEN (64 characters in a standard build).

client_serial numeric Serial number of the client certificate, or NULL if no client certificate was supplied or if SSL is not in use on this connection. The combination of certificate serial number and certificate issuer uniquely identifies a certificate (unless the issuer erroneously reuses serial numbers).

issuer_dn text DN of the issuer of the client certificate, or NULL if no client certificate was supplied or if SSL is not in use on this connection. This field is truncated like client_dn.

5.11.2.11. pg_stat_gssapi

The pg_stat_gssapi view will contain one row per backend, showing information about GSSAPI usage on this connection. It can be joined to pg_stat_activity or pg_stat_replication on the pid column to get more details about the connection.

pg_stat_gssapi View

Column Type Description

pid integer Process ID of a backend

gss_authenticated* * **boolean True if GSSAPI authentication was used for this connection

principal text Principal used to authenticate this connection, or NULL if GSSAPI was not used to authenticate this connection. This field is truncated if the principal is longer than NAMEDATALEN (64 characters in a standard build).

encrypted boolean True if GSSAPI encryption is in use on this connection

5.11.2.12. pg_stat_archiver

The pg_stat_archiver view will always have a single row, containing data about the archiver process of the cluster.

pg_stat_archiver View

Column Type Description

archived_count bigint Number of WAL files that have been successfully archived

last_archived_wal* * **text Name of the WAL file most recently successfully archived

last_archived_time ** **timestamp with time zone Time of the most recent successful archive operation

failed_count bigint Number of failed attempts for archiving WAL files

last_failed_wal text Name of the WAL file of the most recent failed archival operation

last_failed_time timestamp with time zone Time of the most recent failed archival operation

stats_reset timestamp with time zone Time at which these statistics were last reset

Normally, WAL files are archived in order, oldest to newest, but that is not guaranteed, and does not hold under special circumstances like when promoting a standby or after crash recovery. Therefore it is not safe to assume that all files older than last_archived_wal have also been successfully archived.

5.11.2.13. pg_stat_bgwriter

The pg_stat_bgwriter view will always have a single row, containing global data for the cluster.

pg_stat_bgwriter View

Column Type Description

checkpoints_timed* * **bigint Number of scheduled checkpoints that have been performed

checkpoints_req bigint Number of requested checkpoints that have been performed

checkpoint_write_t ime double precision Total amount of time that has been spent in the portion of checkpoint processing where files are written to disk, in milliseconds

checkpoint_sync_ti me double precision Total amount of time that has been spent in the portion of checkpoint processing where files are synchronized to disk, in milliseconds

buffers_checkpoint ** **bigint Number of buffers written during checkpoints

buffers_clean bigint Number of buffers written by the background writer

maxwritten_clean bigint Number of times the background writer stopped a cleaning scan because it had written too many buffers

buffers_backend bigint Number of buffers written directly by a backend

buffers_backend_fs ync bigint Number of times a backend had to execute its own fsync call (normally the background writer handles those even when the backend does its own write)

buffers_alloc bigint Number of buffers allocated

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.14. pg_stat_wal

The pg_stat_wal view will always have a single row, containing data about WAL activity of the cluster.

pg_stat_wal View

Column Type Description

wal_records bigint Total number of WAL records generated

wal_fpi bigint Total number of WAL full page images generated

wal_bytes numeric Total amount of WAL generated in bytes

wal_buffers_full bigint Number of times WAL data was written to disk because WAL buffers became full

wal_write bigint Number of times WAL buffers were written out to disk via XLogWrite request. See wal-configuration for more information about the internal WAL function XLogWrite.

wal_sync bigint Number of times WAL files were synced to disk via issue_xlog_fsync request (if guc-fsync is on and guc-wal-sync-method is either fdatasync, fsync or fsync_writethrough **, otherwise zero). See wal-configuration for more information about the internal WAL function **issue_xlog_fsync .

wal_write_time double precision Total amount of time spent writing WAL buffers to disk via XLogWrite request, in milliseconds (if guc-track-wal-io-tim ing is enabled, otherwise zero). This includes the sync time when wal_sync_method is either open_datasync or open_sync.

wal_sync_time double precision Total amount of time spent syncing WAL files to disk via issue_xlog_fsync request, in milliseconds (if track_wal_io_timin g is enabled, fsync is on, and wal_sync_method is either fdatasync, fsync or **fsync_writethrough **, otherwise zero).

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.15. pg_stat_database

The pg_stat_database view will contain one row for each database in the cluster, plus one for shared objects, showing database-wide statistics.

pg_stat_database View

Column Type Description

datid oid OID of this database, or 0 for objects belonging to a shared relation

datname name Name of this database, or NULL for shared objects.

numbackends integer Number of backends currently connected to this database, or NULL for shared objects. This is the only column in this view that returns a value reflecting current state; all other columns return the accumulated values since the last reset.

xact_commit bigint Number of transactions in this database that have been committed

xact_rollback bigint Number of transactions in this database that have been rolled back

blks_read bigint Number of disk blocks read in this database

blks_hit bigint Number of times disk blocks were found already in the buffer cache, so that a read was not necessary (this only includes hits in the PostgreSQL buffer cache, not the operating system’s file system cache)

tup_returned bigint Number of live rows fetched by sequential scans and index entries returned by index scans in this database

tup_fetched bigint Number of live rows fetched by index scans in this database

tup_inserted bigint Number of rows inserted by queries in this database

tup_updated bigint Number of rows updated by queries in this database

tup_deleted bigint Number of rows deleted by queries in this database

conflicts bigint Number of queries canceled due to conflicts with recovery in this database. (Conflicts occur only on standby servers; see linkend=»monitoring- pg-stat-database-con flicts-view»> pg_stat_database_c onflicts for details.)

temp_files bigint Number of temporary files created by queries in this database. All temporary files are counted, regardless of why the temporary file was created (e.g., sorting or hashing), and regardless of the guc-log-temp-files setting.

temp_bytes bigint Total amount of data written to temporary files by queries in this database. All temporary files are counted, regardless of why the temporary file was created, and regardless of the guc-log-temp-files setting.

deadlocks bigint Number of deadlocks detected in this database

checksum_failures* * **bigint Number of data page checksum failures detected in this database (or on a shared object), or NULL if data checksums are not enabled.

checksum_last_fail ure timestamp with time zone Time at which the last data page checksum failure was detected in this database (or on a shared object), or NULL if data checksums are not enabled.

blk_read_time double precision Time spent reading data file blocks by backends in this database, in milliseconds (if guc-track-io-timing is enabled, otherwise zero)

blk_write_time double precision Time spent writing data file blocks by backends in this database, in milliseconds (if guc-track-io-timing is enabled, otherwise zero)

session_time double precision Time spent by database sessions in this database, in milliseconds (note that statistics are only updated when the state of a session changes, so if sessions have been idle for a long time, this idle time won’t be included)

active_time double precision Time spent executing SQL statements in this database, in milliseconds (this corresponds to the states active and fastpath function call in linkend=»monitoring- pg-stat-activity-vie w»> pg_stat_activity )

idle_in_transactio n_time double precision Time spent idling while in a transaction in this database, in milliseconds (this corresponds to the states idle in transaction and idle in transaction (aborted) in linkend=»monitoring- pg-stat-activity-vie w»> pg_stat_activity )

sessions bigint Total number of sessions established to this database

sessions_abandoned ** **bigint Number of database sessions to this database that were terminated because connection to the client was lost

sessions_fatal bigint Number of database sessions to this database that were terminated by fatal errors

sessions_killed bigint Number of database sessions to this database that were terminated by operator intervention

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.16. pg_stat_database_conflicts

The pg_stat_database_conflicts view will contain one row per database, showing database-wide statistics about query cancels occurring due to conflicts with recovery on standby servers. This view will only contain information on standby servers, since conflicts do not occur on primary servers.

pg_stat_database_conflicts View

Column Type Description

datid oid OID of a database

datname name Name of this database

confl_tablespace bigint Number of queries in this database that have been canceled due to dropped tablespaces

confl_lock bigint Number of queries in this database that have been canceled due to lock timeouts

confl_snapshot bigint Number of queries in this database that have been canceled due to old snapshots

confl_bufferpin bigint Number of queries in this database that have been canceled due to pinned buffers

confl_deadlock bigint Number of queries in this database that have been canceled due to deadlocks

5.11.2.17. pg_stat_all_tables

The pg_stat_all_tables view will contain one row for each table in the current database (including TOAST tables), showing statistics about accesses to that specific table. The pg_stat_user_tables and pg_stat_sys_tables views contain the same information, but filtered to only show user and system tables respectively.

pg_stat_all_tables View

Column Type Description

relid oid OID of a table

schemaname name Name of the schema that this table is in

relname name Name of this table

seq_scan bigint Number of sequential scans initiated on this table

seq_tup_read bigint Number of live rows fetched by sequential scans

idx_scan bigint Number of index scans initiated on this table

idx_tup_fetch bigint Number of live rows fetched by index scans

n_tup_ins bigint Number of rows inserted

n_tup_upd bigint Number of rows updated (includes linkend=»storage-hot «>HOT updated rows)

n_tup_del bigint Number of rows deleted

n_tup_hot_upd bigint Number of rows HOT updated (i.e., with no separate index update required)

n_live_tup bigint Estimated number of live rows

n_dead_tup bigint Estimated number of dead rows

n_mod_since_analyz e bigint Estimated number of rows modified since this table was last analyzed

n_ins_since_vacuum ** **bigint Estimated number of rows inserted since this table was last vacuumed

last_vacuum timestamp with time zone Last time at which this table was manually vacuumed (not counting VACUUM FULL)

last_autovacuum timestamp with time zone Last time at which this table was vacuumed by the autovacuum daemon

last_analyze timestamp with time zone Last time at which this table was manually analyzed

last_autoanalyze timestamp with time zone Last time at which this table was analyzed by the autovacuum daemon

vacuum_count bigint Number of times this table has been manually vacuumed (not counting VACUUM FULL)

autovacuum_count bigint Number of times this table has been vacuumed by the autovacuum daemon

analyze_count bigint Number of times this table has been manually analyzed

autoanalyze_count* * **bigint Number of times this table has been analyzed by the autovacuum daemon

5.11.2.18. pg_stat_all_indexes

The pg_stat_all_indexes view will contain one row for each index in the current database, showing statistics about accesses to that specific index. The pg_stat_user_indexes and pg_stat_sys_indexes views contain the same information, but filtered to only show user and system indexes respectively.

pg_stat_all_indexes View

Column Type Description

relid oid OID of the table for this index

indexrelid oid OID of this index

schemaname name Name of the schema this index is in

relname name Name of the table for this index

indexrelname name Name of this index

idx_scan bigint Number of index scans initiated on this index

idx_tup_read bigint Number of index entries returned by scans on this index

idx_tup_fetch bigint Number of live table rows fetched by simple index scans using this index

Indexes can be used by simple index scans, bitmap index scans, and the optimizer. In a bitmap scan the output of several indexes can be combined via AND or OR rules, so it is difficult to associate individual heap row fetches with specific indexes when a bitmap scan is used. Therefore, a bitmap scan increments the pg_stat_all_indexes.**idx_tup_read** count(s) for the index(es) it uses, and it increments the pg_stat_all_tables.**idx_tup_fetch** count for the table, but it does not affect pg_stat_all_indexes.**idx_tup_fetch**. The optimizer also accesses indexes to check for supplied constants whose values are outside the recorded range of the optimizer statistics because the optimizer statistics might be stale.

Примечание

The idx_tup_read and idx_tup_fetch counts can be different even without any use of bitmap scans, because idx_tup_read counts index entries retrieved from the index while idx_tup_fetch counts live rows fetched from the table. The latter will be less if any dead or not-yet-committed rows are fetched using the index, or if any heap fetches are avoided by means of an index-only scan.

5.11.2.19. pg_statio_all_tables

The pg_statio_all_tables view will contain one row for each table in the current database (including TOAST tables), showing statistics about I/O on that specific table. The pg_statio_user_tables and pg_statio_sys_tables views contain the same information, but filtered to only show user and system tables respectively.

pg_statio_all_tables View

Column Type Description

relid oid OID of a table

schemaname name Name of the schema that this table is in

relname name Name of this table

heap_blks_read bigint Number of disk blocks read from this table

heap_blks_hit bigint Number of buffer hits in this table

idx_blks_read bigint Number of disk blocks read from all indexes on this table

idx_blks_hit bigint Number of buffer hits in all indexes on this table

toast_blks_read bigint Number of disk blocks read from this table’s TOAST table (if any)

toast_blks_hit bigint Number of buffer hits in this table’s TOAST table (if any)

tidx_blks_read bigint Number of disk blocks read from this table’s TOAST table indexes (if any)

tidx_blks_hit bigint Number of buffer hits in this table’s TOAST table indexes (if any)

5.11.2.20. pg_statio_all_indexes

The pg_statio_all_indexes view will contain one row for each index in the current database, showing statistics about I/O on that specific index. The pg_statio_user_indexes and pg_statio_sys_indexes views contain the same information, but filtered to only show user and system indexes respectively.

pg_statio_all_indexes View

Column Type Description

relid oid OID of the table for this index

indexrelid oid OID of this index

schemaname name Name of the schema this index is in

relname name Name of the table for this index

indexrelname name Name of this index

idx_blks_read bigint Number of disk blocks read from this index

idx_blks_hit bigint Number of buffer hits in this index

5.11.2.21. pg_statio_all_sequences

The pg_statio_all_sequences view will contain one row for each sequence in the current database, showing statistics about I/O on that specific sequence.

pg_statio_all_sequences View

Column Type Description

relid oid OID of a sequence

schemaname name Name of the schema this sequence is in

relname name Name of this sequence

blks_read bigint Number of disk blocks read from this sequence

blks_hit bigint Number of buffer hits in this sequence

5.11.2.22. pg_stat_user_functions

The pg_stat_user_functions view will contain one row for each tracked function, showing statistics about executions of that function. The guc-track-functions parameter controls exactly which functions are tracked.

pg_stat_user_functions View

Column Type Description

funcid oid OID of a function

schemaname name Name of the schema this function is in

funcname name Name of this function

calls bigint Number of times this function has been called

total_time double precision Total time spent in this function and all other functions called by it, in milliseconds

self_time double precision Total time spent in this function itself, not including other functions called by it, in milliseconds

5.11.2.23. pg_stat_slru

PostgreSQL accesses certain on-disk information via SLRU (simple least-recently-used) caches. The pg_stat_slru view will contain one row for each tracked SLRU cache, showing statistics about access to cached pages.

pg_stat_slru View

Column Type Description

name text Name of the SLRU

blks_zeroed bigint Number of blocks zeroed during initializations

blks_hit bigint Number of times disk blocks were found already in the SLRU, so that a read was not necessary (this only includes hits in the SLRU, not the operating system’s file system cache)

blks_read bigint Number of disk blocks read for this SLRU

blks_written bigint Number of disk blocks written for this SLRU

blks_exists bigint Number of blocks checked for existence for this SLRU

flushes bigint Number of flushes of dirty data for this SLRU

truncates bigint Number of truncates for this SLRU

stats_reset timestamp with time zone Time at which these statistics were last reset

5.11.2.24. Statistics Functions

Other ways of looking at the statistics can be set up by writing queries that use the same underlying statistics access functions used by the standard views shown above. For details such as the functions“ names, consult the definitions of the standard views. (For example, in psql you could issue d+ pg_stat_activity.) The access functions for per-database statistics take a database OID as an argument to identify which database to report on. The per-table and per-index functions take a table or index OID. The functions for per-function statistics take a function OID. Note that only tables, indexes, and functions in the current database can be seen with these functions.

Additional functions related to the cumulative statistics system are listed in monitoring-stats-funcs-table.

Additional Statistics Functions

Function Description

pg_backend_pid () integer Returns the process ID of the server process attached to the current session.

pg_stat_get_activi ty ( integer ) setof record Returns a record of information about the backend with the specified process ID, or one record for each active backend in the system if NULL is specified. The fields returned are a subset of those in the pg_stat_activity view.

pg_stat_get_snapsh ot_timestamp () timestamp with time zone Returns the timestamp of the current statistics snapshot, or NULL if no statistics snapshot has been taken. A snapshot is taken the first time cumulative statistics are accessed in a transaction if stats_fetch_consis tency is set to snapshot

pg_stat_clear_snap shot () void Discards the current statistics snapshot or cached information.

pg_stat_reset () void Resets all statistics counters for the current database to zero. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_shar ed ( text ) void Resets some cluster-wide statistics counters to zero, depending on the argument. The argument can be bgwriter to reset all the counters shown in the pg_stat_bgwriter view, archiver to reset all the counters shown in the pg_stat_archiver view, wal to reset all the counters shown in the pg_stat_wal view or recovery_prefetch* * to reset all the counters shown in the **pg_stat_recovery_p refetch view. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_sing le_table_counters ( oid ) void Resets statistics for a single table or index in the current database or shared across all databases in the cluster to zero. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_sing le_function_counters ** ( **oid ) void Resets statistics for a single function in the current database to zero. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_slru ** ( **text ) void Resets statistics to zero for a single SLRU cache, or for all SLRUs in the cluster. If the argument is NULL, all counters shown in the pg_stat_slru view for all SLRU caches are reset. The argument can be one of CommitTs, MultiXactMember, MultiXactOffset, Notify, Serial, Subtrans, or Xact to reset the counters for only that entry. If the argument is other (or indeed, any unrecognized name), then the counters for all other SLRU caches, such as extension-defined caches, are reset. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_repl ication_slot ( text ) void Resets statistics of the replication slot defined by the argument. If the argument is NULL, resets statistics for all the replication slots. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

pg_stat_reset_subs cription_stats ( oid ) void Resets statistics for a single subscription shown in the pg_stat_subscripti on_stats view to zero. If the argument is NULL, reset statistics for all subscriptions. This function is restricted to superusers by default, but other users can be granted EXECUTE to run the function.

Предупреждение

Using pg_stat_reset() also resets counters that autovacuum uses to determine when to trigger a vacuum or an analyze. Resetting these counters can cause autovacuum to not perform necessary work, which can cause problems such as table bloat or out-dated table statistics. A database-wide ANALYZE is recommended after the statistics have been reset.

pg_stat_get_activity, the underlying function of the pg_stat_activity view, returns a set of records containing all the available information about each backend process. Sometimes it may be more convenient to obtain just a subset of this information. In such cases, an older set of per-backend statistics access functions can be used; these are shown in monitoring-stats-backend-funcs-table. These access functions use a backend ID number, which ranges from one to the number of currently active backends. The function pg_stat_get_backend_idset provides a convenient way to generate one row for each active backend for invoking these functions. For example, to show the PIDs and current queries of all backends:

SELECT pg_stat_get_backend_pid(s.backendid) AS pid,
       pg_stat_get_backend_activity(s.backendid) AS query
    FROM (SELECT pg_stat_get_backend_idset() AS backendid) AS s;

Per-Backend Statistics Functions

Function Description

pg_stat_get_backen d_idset () setof integer Returns the set of currently active backend ID numbers (from 1 to the number of active backends).

pg_stat_get_backen d_activity ( integer ) text Returns the text of this backend’s most recent query.

pg_stat_get_backen d_activity_start ( integer ) timestamp with time zone Returns the time when the backend’s most recent query was started.

pg_stat_get_backen d_client_addr ( integer ) inet Returns the IP address of the client connected to this backend.

pg_stat_get_backen d_client_port ( integer ) integer Returns the TCP port number that the client is using for communication.

pg_stat_get_backen d_dbid ( integer ) oid Returns the OID of the database this backend is connected to.

pg_stat_get_backen d_pid ( integer ) integer Returns the process ID of this backend.

pg_stat_get_backen d_start ( integer ) timestamp with time zone Returns the time when this process was started.

pg_stat_get_backen d_userid ( integer ) oid Returns the OID of the user logged into this backend.

pg_stat_get_backen d_wait_event_type ( integer ) text Returns the wait event type name if this backend is currently waiting, otherwise NULL. See wait-event-table for details.

pg_stat_get_backen d_wait_event ( integer ) text Returns the wait event name if this backend is currently waiting, otherwise NULL. See wait-event-activity- table through wait-event-timeout-t able.

pg_stat_get_backen d_xact_start ( integer ) timestamp with time zone Returns the time when the backend’s current transaction was started.

5.11.3. Viewing Locks

Another useful tool for monitoring database activity is the pg_locks system table. It allows the database administrator to view information about the outstanding locks in the lock manager. For example, this capability can be used to:

  1. View all the locks currently outstanding, all the locks on relations in a particular database, all the locks on a particular relation, or all the locks held by a particular PostgreSQL session.

  2. Determine the relation in the current database with the most ungranted locks (which might be a source of contention among database clients).

  3. Determine the effect of lock contention on overall database performance, as well as the extent to which contention varies with overall database traffic.

Details of the pg_locks view appear in view-pg-locks. For more information on locking and managing concurrency with PostgreSQL, refer to mvcc.

5.11.4. Progress Reporting

PostgreSQL has the ability to report the progress of certain commands during command execution. Currently, the only commands which support progress reporting are ANALYZE, CLUSTER, CREATE INDEX, VACUUM, COPY, and protocol-replication-base-backup (i.e., replication command that app-pgbasebackup issues to take a base backup). This may be expanded in the future.

5.11.4.1. ANALYZE Progress Reporting

Whenever ANALYZE is running, the pg_stat_progress_analyze view will contain a row for each backend that is currently running that command. The tables below describe the information that will be reported and provide information about how to interpret it.

pg_stat_progress_analyze View

Column Type Description

pid integer Process ID of backend.

datid oid OID of the database to which this backend is connected.

datname name Name of the database to which this backend is connected.

relid oid OID of the table being analyzed.

phase text Current processing phase. See analyze-phases.

sample_blks_total* * **bigint Total number of heap blocks that will be sampled.

sample_blks_scanne d bigint Number of heap blocks scanned.

ext_stats_total bigint Number of extended statistics.

ext_stats_computed ** **bigint Number of extended statistics computed. This counter only advances when the phase is computing extended statistics.

child_tables_total ** **bigint Number of child tables.

child_tables_done* * **bigint Number of child tables scanned. This counter only advances when the phase is acquiring inherited sample rows.

current_child_tabl e_relid oid OID of the child table currently being scanned. This field is only valid when the phase is acquiring inherited sample rows.

ANALYZE Phases

Примечание

Note that when ANALYZE is run on a partitioned table, all of its partitions are also recursively analyzed. In that case, ANALYZE progress is reported first for the parent table, whereby its inheritance statistics are collected, followed by that for each partition.

5.11.4.2. CREATE INDEX Progress Reporting

Whenever CREATE INDEX or REINDEX is running, the pg_stat_progress_create_index view will contain one row for each backend that is currently creating indexes. The tables below describe the information that will be reported and provide information about how to interpret it.

pg_stat_progress_create_index View

Column Type Description

pid integer Process ID of backend.

datid oid OID of the database to which this backend is connected.

datname name Name of the database to which this backend is connected.

relid oid OID of the table on which the index is being created.

index_relid oid OID of the index being created or reindexed. During a non-concurrent CREATE INDEX, this is 0.

command text The command that is running: CREATE INDEX, CREATE INDEX CONCURRENTLY, REINDEX, or REINDEX CONCURRENTLY.

phase text Current processing phase of index creation. See create-index-phases.

lockers_total bigint Total number of lockers to wait for, when applicable.

lockers_done bigint Number of lockers already waited for.

current_locker_pid ** **bigint Process ID of the locker currently being waited for.

blocks_total bigint Total number of blocks to be processed in the current phase.

blocks_done bigint Number of blocks already processed in the current phase.

tuples_total bigint Total number of tuples to be processed in the current phase.

tuples_done bigint Number of tuples already processed in the current phase.

partitions_total bigint When creating an index on a partitioned table, this column is set to the total number of partitions on which the index is to be created. This field is 0 during a REINDEX.

partitions_done bigint When creating an index on a partitioned table, this column is set to the number of partitions on which the index has been created. This field is 0 during a REINDEX.

CREATE INDEX Phases

5.11.4.3. VACUUM Progress Reporting

Whenever VACUUM is running, the pg_stat_progress_vacuum view will contain one row for each backend (including autovacuum worker processes) that is currently vacuuming. The tables below describe the information that will be reported and provide information about how to interpret it. Progress for VACUUM FULL commands is reported via pg_stat_progress_cluster because both VACUUM FULL and CLUSTER rewrite the table, while regular VACUUM only modifies it in place. See cluster-progress-reporting.

pg_stat_progress_vacuum View

Column Type Description

pid integer Process ID of backend.

datid oid OID of the database to which this backend is connected.

datname name Name of the database to which this backend is connected.

relid oid OID of the table being vacuumed.

phase text Current processing phase of vacuum. See vacuum-phases.

heap_blks_total bigint Total number of heap blocks in the table. This number is reported as of the beginning of the scan; blocks added later will not be (and need not be) visited by this VACUUM.

heap_blks_scanned* * **bigint Number of heap blocks scanned. Because the linkend=»storage-vm» >visibility map is used to optimize scans, some blocks will be skipped without inspection; skipped blocks are included in this total, so that this number will eventually become equal to heap_blks_total when the vacuum is complete. This counter only advances when the phase is scanning heap.

heap_blks_vacuumed ** **bigint Number of heap blocks vacuumed. Unless the table has no indexes, this counter only advances when the phase is vacuuming heap. Blocks that contain no dead tuples are skipped, so the counter may sometimes skip forward in large increments.

index_vacuum_count ** **bigint Number of completed index vacuum cycles.

max_dead_tuples bigint Number of dead tuples that we can store before needing to perform an index vacuum cycle, based on guc-maintenance-work -mem.

num_dead_tuples bigint Number of dead tuples collected since the last index vacuum cycle.

VACUUM Phases

5.11.4.4. CLUSTER Progress Reporting

Whenever CLUSTER or VACUUM FULL is running, the pg_stat_progress_cluster view will contain a row for each backend that is currently running either command. The tables below describe the information that will be reported and provide information about how to interpret it.

pg_stat_progress_cluster View

Column Type Description

pid integer Process ID of backend.

datid oid OID of the database to which this backend is connected.

datname name Name of the database to which this backend is connected.

relid oid OID of the table being clustered.

command text The command that is running. Either CLUSTER or VACUUM FULL.

phase text Current processing phase. See cluster-phases.

cluster_index_reli d oid If the table is being scanned using an index, this is the OID of the index being used; otherwise, it is zero.

heap_tuples_scanne d bigint Number of heap tuples scanned. This counter only advances when the phase is seq scanning heap, index scanning heap or writing new heap.

heap_tuples_writte n bigint Number of heap tuples written. This counter only advances when the phase is seq scanning heap, index scanning heap or writing new heap.

heap_blks_total bigint Total number of heap blocks in the table. This number is reported as of the beginning of seq scanning heap.

heap_blks_scanned* * **bigint Number of heap blocks scanned. This counter only advances when the phase is seq scanning heap.

index_rebuild_coun t bigint Number of indexes rebuilt. This counter only advances when the phase is rebuilding index.

CLUSTER and VACUUM FULL Phases

5.11.4.5. Base Backup Progress Reporting

Whenever an application like pg_basebackup is taking a base backup, the pg_stat_progress_basebackup view will contain a row for each WAL sender process that is currently running the BASE_BACKUP replication command and streaming the backup. The tables below describe the information that will be reported and provide information about how to interpret it.

pg_stat_progress_basebackup View

Column Type Description

pid integer Process ID of a WAL sender process.

phase text Current processing phase. See basebackup-phases.

backup_total bigint Total amount of data that will be streamed. This is estimated and reported as of the beginning of streaming database files phase. Note that this is only an approximation since the database may change during streaming database files phase and WAL log may be included in the backup later. This is always the same value as backup_streamed once the amount of data streamed exceeds the estimated total size. If the estimation is disabled in pg_basebackup (i.e., –no-estimate-size ** option is specified), this is **NULL.

backup_streamed bigint Amount of data streamed. This counter only advances when the phase is streaming database files or transferring wal files.

tablespaces_total* * **bigint Total number of tablespaces that will be streamed.

tablespaces_stream ed bigint Number of tablespaces streamed. This counter only advances when the phase is streaming database files.

Base Backup Phases

5.11.4.6. COPY Progress Reporting

Whenever COPY is running, the pg_stat_progress_copy view will contain one row for each backend that is currently running a COPY command. The table below describes the information that will be reported and provides information about how to interpret it.

pg_stat_progress_copy View

Column Type Description

pid integer Process ID of backend.

datid oid OID of the database to which this backend is connected.

datname name Name of the database to which this backend is connected.

relid oid OID of the table on which the COPY command is executed. It is set to 0 if copying from a SELECT query.

command text The command that is running: COPY FROM, or COPY TO.

type text The io type that the data is read from or written to: FILE, PROGRAM, PIPE (for COPY FROM STDIN and COPY TO STDOUT), or CALLBACK (used for example during the initial table synchronization in logical replication).

bytes_processed bigint Number of bytes already processed by COPY command.

bytes_total bigint Size of source file for COPY FROM command in bytes. It is set to 0 if not available.

tuples_processed bigint Number of tuples already processed by COPY command.

tuples_excluded bigint Number of tuples not processed because they were excluded by the WHERE clause of the COPY command.

5.11.5. Dynamic Tracing

PostgreSQL provides facilities to support dynamic tracing of the database server. This allows an external utility to be called at specific points in the code and thereby trace execution.

A number of probes or trace points are already inserted into the source code. These probes are intended to be used by database developers and administrators. By default the probes are not compiled into PostgreSQL; the user needs to explicitly tell the configure script to make the probes available.

Currently, the DTrace utility is supported, which, at the time of this writing, is available on Solaris, macOS, FreeBSD, NetBSD, and Oracle Linux. The SystemTap project for Linux provides a DTrace equivalent and can also be used. Supporting other dynamic tracing utilities is theoretically possible by changing the definitions for the macros in src/include/utils/probes.h.

5.11.5.1. Compiling for Dynamic Tracing

By default, probes are not available, so you will need to explicitly tell the configure script to make the probes available in PostgreSQL. To include DTrace support specify –enable-dtrace to configure. See install-procedure for further information.

5.11.5.2. Built-in Probes

A number of standard probes are provided in the source code, as shown in dtrace-probe-point-table; typedefs-table shows the types used in the probes. More probes can certainly be added to enhance PostgreSQL’s observability.

Built-in DTrace Probes

Defined Types Used in Probe Parameters

5.11.5.3. Using Probes

The example below shows a DTrace script for analyzing transaction counts in the system, as an alternative to snapshotting pg_stat_database before and after a performance test:

#!/usr/sbin/dtrace -qs

postgresql$1:::transaction-start
{
      @start["Start"] = count();
      self->ts  = timestamp;
}

postgresql$1:::transaction-abort
{
      @abort["Abort"] = count();
}

postgresql$1:::transaction-commit
/self->ts/
{
      @commit["Commit"] = count();
      @time["Total time (ns)"] = sum(timestamp - self->ts);
      self->ts=0;
}
 When executed, the example D script gives output such as:
# ./txn_count.d `pgrep -n postgres` or ./txn_count.d <PID>
^C

Start                                          71
Commit                                         70
Total time (ns)                        2312105013

Примечание

SystemTap uses a different notation for trace scripts than DTrace does, even though the underlying trace points are compatible. One point worth noting is that at this writing, SystemTap scripts must reference probe names using double underscores in place of hyphens. This is expected to be fixed in future SystemTap releases.

You should remember that DTrace scripts need to be carefully written and debugged, otherwise the trace information collected might be meaningless. In most cases where problems are found it is the instrumentation that is at fault, not the underlying system. When discussing information found using dynamic tracing, be sure to enclose the script used to allow that too to be checked and discussed.

5.11.5.4. Defining New Probes

New probes can be defined within the code wherever the developer desires, though this will require a recompilation. Below are the steps for inserting new probes:

Decide on probe names and data to be made available through the probes

Add the probe definitions to src/backend/utils/probes.d

Include pg_trace.h if it is not already present in the module(s) containing the probe points, and insert TRACE_POSTGRESQL probe macros at the desired locations in the source code

Recompile and verify that the new probes are available

5.11.5.4.1. Example:

Here is an example of how you would add a probe to trace all new transactions by transaction ID.

Decide that the probe will be named transaction-start and requires a parameter of type LocalTransactionId

Add the probe definition to src/backend/utils/probes.d:

probe transaction__start(LocalTransactionId);
   Note the use of the double underline in the probe name. In a DTrace

script using the probe, the double underline needs to be replaced with a hyphen, so transaction-start is the name to document for users.

At compile time, transaction__start is converted to a macro called TRACE_POSTGRESQL_TRANSACTION_START (notice the underscores are single here), which is available by including pg_trace.h. Add the macro call to the appropriate location in the source code. In this case, it looks like the following:

TRACE_POSTGRESQL_TRANSACTION_START(vxid.localTransactionId);

After recompiling and running the new binary, check that your newly added probe is available by executing the following DTrace command. You should see similar output:

# dtrace -ln transaction-start
   ID    PROVIDER          MODULE           FUNCTION NAME
18705 postgresql49878     postgres     StartTransactionCommand transaction-start
18755 postgresql49877     postgres     StartTransactionCommand transaction-start
18805 postgresql49876     postgres     StartTransactionCommand transaction-start
18855 postgresql49875     postgres     StartTransactionCommand transaction-start
18986 postgresql49873     postgres     StartTransactionCommand transaction-start

There are a few things to be careful about when adding trace macros to the C code:

  1. You should take care that the data types specified for a probe’s parameters match the data types of the variables used in the macro. Otherwise, you will get compilation errors.

  2. On most platforms, if PostgreSQL is built with –enable-dtrace, the arguments to a trace macro will be evaluated whenever control passes through the macro, even if no tracing is being done. This is usually not worth worrying about if you are just reporting the values of a few local variables. But beware of putting expensive function calls into the arguments. If you need to do that, consider protecting the macro with a check to see if the trace is actually enabled:

    if (TRACE_POSTGRESQL_TRANSACTION_START_ENABLED())
        TRACE_POSTGRESQL_TRANSACTION_START(some_function(...));
    

    Each trace macro has a corresponding ENABLED macro.