Instances - PolarDB - Alibaba Cloud Documentation Center

This topic describes the views in PolarDB for PostgreSQL. You can query views to obtain instance details. This helps you troubleshoot performance issues.

Basic instance information

pg_stat_database

The pg_stat_database view displays the statistics about each row in each instance of a cluster. The following table describes the parameters in this view.


Parameter	Type	Description
`datid`	oid	The object identifier (OID) of the database.
`datname`	name	The database name.
`numbackends`	integer	The number of backend servers that are connected to the database. In the view, this column is the only column that returns the current status. The other columns return the cumulative values starting from the last value reset.
`xact_commit`	bigint	The total number of transactions that have been committed in the database.
`xact_rollback`	bigint	The total number of transactions that have been rolled back in the database.
`blks_read`	bigint	The total number of disk blocks that have been physically read in the database.
`blks_hit`	bigint	The number of disk blocks that have been hit in the buffer. The disk blocks do not need to be read. The number of cache hits indicates only the hits in the PolarDB for PostgreSQL buffer. The hits in the file system buffer of the OS are ignored.
`tup_returned`	bigint	The number of rows returned by queries in the database.
`tup_fetched`	bigint	The number of rows fetched by queries in the database.
`tup_inserted`	bigint	The number of rows inserted by queries in the database.
`tup_updated`	bigint	The number of rows updated by queries in the database.
`tup_deleted`	bigint	The number of rows deleted by queries in the database.
`conflicts`	bigint	The number of queries that are canceled due to the recovery of read-only node replay conflicts.
`temp_files`	bigint	The number of temporary files created by queries in the database. All temporary files are counted regardless of why the files are created and regardless of what the `log_temp_files` settings are.
`temp_bytes`	bigint	The total amount of data written to temporary files by queries in the database. All temporary files are counted regardless of why the files are created and regardless of what the `log_temp_files` settings are.
`deadlocks`	bigint	The number of deadlocks detected in the database.
`blk_read_time`	double precision	The amount of time spent by backend servers for reading data files in the database. (Unit: milliseconds)
`blk_write_time`	double precision	The time spent by backend servers for writing data files in the database. (Unit: milliseconds)
`stats_reset`	timestamp with time zone	The latest time when the statistics were reset.

pg_stat_bgwriter

The pg_stat_bgwriter view contains only one row. This row contains the global data of the cluster.


Parameter	Type	Description
`checkpoints_timed`	bigint	The number of checkpoints that have been performed.
`checkpoints_req`	bigint	The number of requested checkpoints that have been performed.
`checkpoint_write_time`	double precision	The amount of time spent in writing the file to disks during checkpoints. (Unit: milliseconds)
`checkpoint_sync_time`	double precision	The amount of time spent in synchronizing the file to disks during checkpoints. (Unit: milliseconds)
`buffers_checkpoint`	bigint	The number of buffers that are written during checkpoints.
`buffers_clean`	bigint	The number of buffers that are occupied by the background writer process.
`maxwritten_clean`	bigint	The number of times that the background writer process stops a cleaning scan because a large volume of data is written at a time.
`buffers_backend`	bigint	The number of buffers that are directly written by a backend.
`buffers_backend_fsync`	bigint	The number of times that a backend had executed the fsync call instead of a background writer process. In most cases, the background writer process handles write operations even when the backend also performs the write operations.
`buffers_alloc`	bigint	The number of allocated buffers.
`stats_reset`	timestamp with time zone	The last time when the statistics were reset.

Activity status

polar_stat_activity

The polar_stat_activity view displays the status of each process. It provides the cumulative sum for each process.


Parameter	Type	Description
`datid`	oid	The OID of the database.
`datname`	name	The database name.
`pid`	integer	The ID of the process.
`usesysid`	oid	The ID of the user.
`usename`	name	The user name.
`application_name`	text	The name of the application that is connected to the database.
`client_addr`	inet	The IP address of the client that is connected to the database.
`client_hostname`	text	The host name of the connected client.
`client_port`	integer	The port that is used for the communications between the client and the database.
`backend_start`	timestamp	The time when the process is started.
`xact_start`	timestamp	The time when the current transaction of the process is started.
`query_start`	timestamp	The start time when the SQL statement is executed.
`state_change`	timestamp	The time when the status of the process is last changed.
`wait_event_type`	text	The type of the event for which the backend is waiting in a session. If the backend is not waiting for an event, NULL is returned. Valid values: `LWLock`: The backend is waiting for a lightweight lock. A lightweight lock protects a special data structure in shared memory. The `wait_event` parameter contains a name that identifies the purpose of the lightweight lock. Some locks have specific names. Other locks are a group of locks that have a similar purpose. `Lock`: The backend is waiting for a heavyweight lock. A heavyweight lock is also known as a lock manager or a simple lock. A heavyweight lock protects SQL objects such as tables. The SQL objects are visible to users. A heavyweight lock is also used to ensure mutual exclusion for some internal operations such as table extension. The `wait_event` parameter indicates the type of the lock for which the backend is waiting. `BufferPin`: The server process is waiting to access a data buffer during a period when no other process is checking that buffer. Buffer pin waits can be protracted if another process holds an open cursor that last reads data from the buffer. `Activity`: The server process is idle. This parameter is used when a system process waits for an activity in the main processing loop. `wait_event` identifies the specific wait point. `Extension`: The server process is waiting for an activity in an extension module. This category is useful for modules to track custom wait points. `Client`: The server process is waiting for an activity on a socket from user applications. The server also expects events that are independent of the internal processes to happen. `wait_event` identifies the specific wait point. `IPC`: The server process is waiting for an activity from another process in the server. `wait_event` identifies the specific wait point. `Timeout`: The server process is waiting for a timeout. `wait_event` identifies the specific wait point. `IO`: The server process is waiting for an I/O request to complete. `wait_event` identifies the specific wait point.
`wait_event`	text	The name of the wait event. If the backend is not waiting for an event, NULL is returned.
`state`	text	The status of the process.
`backend_xid`	xid	The transaction identifier (XID) of the client.
`backend_xmin`	xid	The xmin value of the client.
`query`	text	The executed SQL statement.
`backend_type`	text	The type of the backend process.
`queryid`	bigint	The unique ID of the SQL statement.
`wait_object`	text	The wait object. This parameter is returned when the backend is waiting for an event.
`wait_type`	text	The type of the wait object. This parameter is returned when the backend is waiting for an event.
`wait_time_ms`	double	The wait time that is spent. This parameter is returned when the backend is waiting for an event.
`cpu_user`	bigint	The user CPU utilization of the backend server.
`cpu_sys`	bigint	The system CPU utilization of the backend server.
`rss`	bigint	The memory usage of the backend server.
`pfs_read_ps`	bigint	The cumulative sum of read operations on a Polar file system (PFS).
`pfs_write_ps`	bigint	The cumulative sum of write operations on a PFS.
`pfs_read_throughput`	bigint	The cumulative sum of the read throughput of a PFS.
`pfs_write_throughput`	bigint	The cumulative sum of the write throughput of a PFS.
`pfs_read_latency_ms`	double	The latency of the read operation on a PFS.
`pfs_write_latency_ms`	double	The latency of the write operation on a PFS.
`local_read_ps`	bigint	The cumulative sum of read operations on a local file system.
`local_write_ps`	bigint	The cumulative sum of write operations on a local file system.
`local_read_throughput`	bigint	The cumulative sum of the read throughput of a local file system.
`local_write_throughput`	bigint	The cumulative sum of the write throughput of a local file system.
`local_read_latency_ms`	double	The latency of the read operation on a local file system.
`local_write_latency_ms`	double	The latency of the write operation on a local file system.

Wait events


Wait event type	Wait event name	Description
`LWLock`	ShmemIndexLock	The program is waiting to find or allocate space in shared memory.
	OidGenLock	The program is waiting to allocate or assign an OID.
	XidGenLock	The program is waiting to allocate or assign an XID.
	ProcArrayLock	The program is waiting to obtain a snapshot or clear an XID at the end of a transaction.
	SInvalReadLock	The program is waiting to retrieve or remove a message from an invalid shared message queue.
	SInvalWriteLock	The program is waiting to add a message to an invalid shared message queue.
	WALBufMappingLock	The program is waiting to replace a page in a write-ahead logging (WAL) buffer.
	WALWriteLock	The program is waiting for a WAL buffer to be written to a disk.
	ControlFileLock	The program is waiting to read or update the control file or create a WAL file.
	CheckpointLock	The program is waiting to perform a checkpoint operation.
	CLogControlLock	The program is waiting to read or update the transaction status.
	SubtransControlLock	The program is waiting to read or update sub-transaction information.
	MultiXactGenLock	The program is waiting to read or update the shared multixact status.
	MultiXactOffsetControlLock	The program is waiting to read or update multixact offset mappings.
	MultiXactMemberControlLock	The program is waiting to read or update multixact member mappings.
	RelCacheInitLock	The program is waiting to read or write a relation-cache initialization file.
	CheckpointerCommLock	The program is waiting to manage fsync() calls.
	TwoPhaseStateLock	The program is waiting to read or update the status of a prepared transaction.
	TablespaceCreateLock	The program is waiting to create or drop a tablespace.
	BtreeVacuumLock	The program is waiting to read or update vacuum-related information for a B-tree index.
	AddinShmemInitLock	The program is waiting to manage space allocation in shared memory.
	AutovacuumLock	The autovacuum worker or launcher is waiting to update or read the status of the autovacuum worker.
	AutovacuumScheduleLock	The program is waiting to confirm that the table selected for a vacuum still needs vacuuming.
	SyncScanLock	The program is waiting to obtain the start location of a scan on a table for synchronized scans.
	RelationMappingLock	The program is waiting to update the relation mapping file used to store the catalog to filenode mappings.
	AsyncCtlLock	The program is waiting to read or update the shared notification status.
	AsyncQueueLock	The program is waiting to read or update notification messages.
	SerializableXactHashLock	The program is waiting to retrieve or store information about serializable transactions.
	SerializableFinishedListLock	The program is waiting to access finished serializable transactions.
	SerializablePredicateLockListLock	The program is waiting to perform an operation on locks held by serializable transactions.
	OldSerXidLock	The program is waiting to read or record conflicting serializable transactions.
	SyncRepLock	The program is waiting to read or update information about synchronous replicas.
	BackgroundWorkerLock	The program is waiting to read or update the backend worker status.
	DynamicSharedMemoryControlLock	The program is waiting to read or update the status of the dynamic shared memory.
	AutoFileLock	The program is waiting to update the postgresql.auto.conf file.
	ReplicationSlotAllocationLock	The program is waiting to allocate or drop a replication slot.
	ReplicationSlotControlLock	The program is waiting to read or update the status of a replication slot.
	CommitTsControlLock	The program is waiting to read or update transaction commit timestamps.
	CommitTsLock	The program is waiting to read or update the last value set for the transaction timestamp.
	ReplicationOriginLock	The program is waiting to set up, drop, or use the replication origin.
	MultiXactTruncationLock	The program is waiting to read or truncate multixact information.
	OldSnapshotTimeMapLock	The program is waiting to read or update control information about old snapshots.
	BackendRandomLock	The program is waiting to generate a random number.
	LogicalRepWorkerLock	The program is waiting for an action on a logical replication worker to complete.
	CLogTruncationLock	The program is waiting to truncate a WAL file or waiting for the truncation of a WAL file to complete.
	clog	The backend is waiting for I/O requests on a commit log (CLOG) buffer. A CLOG records the commit status of a transaction.
	commit_timestamp	The program is waiting for I/O requests on a commit timestamp buffer.
	subtrans	The program is waiting for I/O requests on a subtransaction buffer.
	multixact_offset	The program is waiting for I/O requests on a multixact offset buffer.
	multixact_member	The program is waiting for I/O requests on a multixact member buffer.
	async	The program is waiting for I/O requests on an async buffer.
	oldserxid	The program is waiting for I/O requests on an oldserxid buffer.
	wal_insert	The program is waiting to insert a WAL file into a memory buffer.
	buffer_content	The program is waiting to read or write a data page in memory.
	buffer_io	The program is waiting for I/O requests on a data page.
	replication_origin	The program is waiting to read or update the replication progress.
	replication_slot_io	The program is waiting for I/O requests on a replication slot.
	proc	The program is waiting to read or update the fast-path lock information.
	buffer_mapping	The program is waiting to associate a data block with a buffer in a buffer pool.
	lock_manager	The program is waiting to join or exit a locking group, or waiting to add or check the locks that are used for the backend. The event takes place when parallel queries are performed.
	predicate_lock_manager	The program is waiting to add or check information about a predicate lock.
	parallel_query_dsa	The program is waiting for the dynamic shared memory allocation lock for parallel execution.
	tbm	The program is waiting for the shared iterator lock for a TIDBitmap.
	parallel_append	The program is waiting to select the next subplan while the Parallel Append plan is being executed.
parallel_hash_join	The program is waiting to allocate or exchange a chunk of memory or update counters while the Parallel Hash plan is being executed.
`Lock`	relation	The program is waiting to obtain the lock on a relation.
	extend	The program is waiting to extend a relation.
	page	The program is waiting to obtain the lock on the page of a relation.
	tuple	The program is waiting to obtain the lock on a tuple.
	transactionid	The program is waiting for a transaction to complete.
	virtualxid	The program is waiting to obtain a virtual XID lock.
	speculative token	The program is waiting to obtain a speculative insertion lock.
	object	The program is waiting to obtain the lock on a non-relational database object.
	userlock	The program is waiting to obtain a user lock.
	advisory	The program is waiting to obtain an advisory user lock.
`BufferPin`	BufferPin	The program is waiting to pin a buffer.
`Activity`	ArchiverMain	The program is waiting in the main loop of the archiver process.
	AutoVacuumMain	The program is waiting in the main loop of the autovacuum launcher process.
	BgWriterHibernate	The program is waiting in the background writer process while the instance is in hibernation.
	BgWriterMain	The program is waiting in the main loop of the background worker of the background worker process.
	CheckpointerMain	The program is waiting in the main loop of the checkpointer process.
	LogicalApplyMain	The program is waiting in the main loop of the logical apply process.
	LogicalLauncherMain	The program is waiting in the main loop of the logical launcher process.
	PgStatMain	The program is waiting in the main loop of the statistics collector process.
	RecoveryWalAll	The program is waiting for the WAL data from a source, such as a local file, an archived file, or a stream, while the instance is being restored.
	RecoveryWalStream	The program is waiting for the WAL data from a stream while the instance is being restored.
	SysLoggerMain	The program is waiting in the main loop of the syslog process.
	WalReceiverMain	The program is waiting in the main loop of the WAL receiver process.
	WalSenderMain	The program is waiting in the main loop of the WAL sender process.
	WalWriterMain	The program is waiting in the main loop of the WAL writer process.
`Client`	ClientRead	The program is waiting to read data from the client.
	ClientWrite	The program is waiting to write data to the client.
	LibPQWalReceiverConnect	The program is waiting in the WAL receiver process to connect to a remote server.
	LibPQWalReceiverReceive	The program is waiting in the WAL receiver process to receive data from a remote server.
	SSLOpenServer	The program is waiting for SSL while a connection is being established.
	WalReceiverWaitStart	The program is waiting for the startup process to send the initial data for streaming replication.
	WalSenderWaitForWAL	The program is waiting for the WAL data to be flushed in the WAL sender process.
	WalSenderWriteData	The program is waiting for an activity that occurs while the replies from the WAL receiver are being processed in the WAL sender process.
`Extension`	Extension	The program is waiting in an extension.
`IPC`	BgWorkerShutdown	The program is waiting for the background worker to shut down.
	BgWorkerStartup	The program is waiting for the background worker to start up.
	BtreePage	The program is waiting for the page number that is required for continuing to scan B-trees in parallel to become available.
	ClogGroupUpdate	The program is waiting for the group leader to update the status of a transaction when the transaction ends.
	ExecuteGather	The program is waiting for an activity from a child process while a Gather node is being executed.
	Hash/Batch/Allocating	The program is waiting for an elected Parallel Hash participant to allocate hash tables.
	Hash/Batch/Electing	The program is waiting for a Parallel Hash participant to be elected to allocate hash tables.
	Hash/Batch/Loading	The program is waiting for other Parallel Hash participants to finish loading hash tables.
	Hash/Build/Allocating	The program is waiting for an elected Parallel Hash participant to allocate the initial hash table.
	Hash/Build/Electing	The program is waiting for a Parallel Hash participant to be elected to allocate the initial hash table.
	Hash/Build/HashingInner	The program is waiting for other Parallel Hash participants to complete the hash operations on inner relations.
	Hash/Build/HashingOuter	The program is waiting for other Parallel Hash participants to complete the hash operations on outer relations.
	Hash/GrowBatches/Allocating	The program is waiting for an elected Parallel Hash participant to allocate more batches of hash tables.
	Hash/GrowBatches/Deciding	The program is waiting for a Parallel Hash participant to be elected to determine the increase of hash table batches.
	Hash/GrowBatches/Electing	The program is waiting for a Parallel Hash participant to be elected to allocate more batches of hash tables.
	Hash/GrowBatches/Finishing	The program is waiting for an elected Parallel Hash participant to determine the increase of hash table batches.
	Hash/GrowBatches/Repartitioning	The program is waiting for other Parallel Hash participants to finish repartitioning.
	Hash/GrowBuckets/Allocating	The program is waiting for an elected Parallel Hash participant to finish allocating more buckets.
	Hash/GrowBuckets/Electing	The program is waiting for a Parallel Hash participant to be elected to allocate more buckets.
	Hash/GrowBuckets/Reinserting	The program is waiting for other Parallel Hash participants to insert tuples into new buckets.
	LogicalSyncData	The program is waiting for the remote server for logical replication to send data that is used to synchronize the initial table.
	LogicalSyncStateChange	The program is waiting for the status of the remote server for logical replication to change.
	MessageQueueInternal	The program is waiting for other processes to be added into a shared message queue.
	MessageQueuePutMessage	The program is waiting to write a protocol message to a shared message queue.
	MessageQueueReceive	The program is waiting to receive bytes from a shared message queue.
	MessageQueueSend	The program is waiting to send bytes to a shared message queue.
	ParallelBitmapScan	The program is waiting for the parallel scan of a bitmap to be initialized.
	ParallelCreateIndexScan	The program is waiting for CREATE INDEX parallel workers to complete heap scans.
	ParallelFinish	The program is waiting for parallel workers to finish computing.
	ProcArrayGroupUpdate	The program is waiting for the group leader to clear the XID after a transaction is complete.
	ReplicationOriginDrop	The program is waiting for a replication origin to become inactive so that the replication origin can be deleted.
	ReplicationSlotDrop	The program is waiting for a replication slot to become inactive so that the replication slot can be deleted.
	SafeSnapshot	The program is waiting for a snapshot that can be used in a READ ONLY DEFERRABLE transaction.
	SyncRep	The program is waiting for a confirmation from the remote server during the synchronous replication.
`Timeout`	BaseBackupThrottle	The program is waiting in the base backup phase while a throttling activity is in progress.
	PgSleep	The program is waiting in the process that calls the pg_sleep function.
	RecoveryApplyDelay	The program is waiting to apply WAL while data is being restored because WAL is delayed to be applied.
`IO`	BufFileRead	The program is waiting to read data from a buffered file.
	BufFileWrite	The program is waiting to write data to a buffered file.
	ControlFileRead	The program is waiting to read data from the control file.
	ControlFileSync	The program is waiting for the control file to reach stable storage.
	ControlFileSyncUpdate	The program is waiting for an update to the control file to reach stable storage.
	ControlFileWrite	The program is waiting to write data to the control file.
	ControlFileWriteUpdate	The program is waiting to write data to update the control file.
	CopyFileRead	The program is waiting to read data while a file is being copied.
	CopyFileWrite	The program is waiting to write data while a file is being copied.
	DataFileExtend	The program is waiting for a relation data file to be extended.
	DataFileFlush	The program is waiting for a relation data file to reach stable storage.
	DataFileImmediateSync	The program is waiting for a relation data file to be immediately synchronized to reach stable storage.
	DataFilePrefetch	The program is waiting for an asynchronous prefetch from a relation data file.
	DataFileRead	The program is waiting to read data from a relation data file.
	DataFileSync	The program is waiting for changes to a relation data file to reach stable storage.
	DataFileTruncate	The program is waiting for a relation data file to be truncated.
	DataFileWrite	The program is waiting to write data to a relation data file.
	DSMFillZeroWrite	The program is waiting to write 0 bytes to a backup file in dynamic shared memory.
	LockFileAddToDataDirRead	The program is waiting to read data while a row of data is being added to the data directory lock file.
	LockFileAddToDataDirSync	The program is waiting for a row of data to reach stable storage while the row of data is being added to the data directory lock file.
	LockFileAddToDataDirWrite	The program is waiting to write data while a row of data is being added to the data directory lock file.
	LockFileCreateRead	The program is waiting to read data while the data directory lock file is being created.
	LockFileCreateSync	The program is waiting for data to reach stable storage while the data directory lock file is being created.
	LockFileCreateWrite	The program is waiting to write data while the data directory lock file is being created.
	LockFileReCheckDataDirRead	The program is waiting to read data while the data directory lock file is being rechecked.
	LogicalRewriteCheckpointSync	The program is waiting for logical rewrite mappings to reach stable storage during a checkpoint.
	LogicalRewriteMappingSync	The program is waiting for mapping data to reach stable storage during a logical rewrite operation.
	LogicalRewriteMappingWrite	The program is waiting to write mapping data during a logical rewrite operation.
	LogicalRewriteSync	The program is waiting for logical rewrite mappings to reach stable storage.
	LogicalRewriteWrite	The program is waiting to write data to logical rewrite mappings.
	RelationMapRead	The program is waiting to read data from the relation mapping file.
	RelationMapSync	The program is waiting for the relation mapping file to reach stable storage.
	RelationMapWrite	The program is waiting to write data to the relation mapping file.
	ReorderBufferRead	The program is waiting to read data during reorder buffer management.
	ReorderBufferWrite	The program is waiting to write data during reorder buffer management.
	ReorderLogicalMappingRead	The program is waiting to read data from a logical mapping during reorder buffer management.
	ReplicationSlotRead	The program is waiting to read data from the control file of a replication slot.
	ReplicationSlotRestoreSync	The program is waiting for the control file of a replication slot to reach stable storage while the file is being restored to the memory.
	ReplicationSlotSync	The program is waiting for the control file of a replication slot to reach stable storage.
	ReplicationSlotWrite	The program is waiting to write data to the control file of a replication slot.
	SLRUFlushSync	The program is waiting for Segmented Least Recently Used (SLRU) data to reach stable storage during a checkpoint or database shutdown period.
	SLRURead	The program is waiting to read data from an SLRU page.
	SLRUSync	The program is waiting for SLRU data to reach stable storage after a write operation is performed on a page.
	SLRUWrite	The program is waiting to write data to an SLRU page.
	SnapbuildRead	The program is waiting to read data from a serialized historical catalog snapshot.
	SnapbuildSync	The program is waiting for a serialized historical catalog snapshot to reach stable storage.
	SnapbuildWrite	The program is waiting to write data to a serialized historical catalog snapshot.
	TimelineHistoryFileSync	The program is waiting for a timeline history file that is received by using streaming replication to reach stable storage.
	TimelineHistoryFileWrite	The program is waiting to read data from a timeline history file that is received by using streaming replication.
	TimelineHistoryRead	The program is waiting to read data from a timeline history file.
	TimelineHistorySync	The program is waiting for a newly created timeline history file to reach stable storage.
	TimelineHistoryWrite	The program is waiting to write data to a newly created timeline history file.
	TwophaseFileRead	The program is waiting to read data from a two-phase state file.
	TwophaseFileSync	The program is waiting for a two-phase state file to reach stable storage.
	TwophaseFileWrite	The program is waiting to write data to a two-phase state file.
	WALBootstrapSync	The program is waiting for WAL data to reach stable storage during bootstrapping.
	WALBootstrapWrite	The program is waiting to write data to a WAL page during bootstrapping.
	WALCopyRead	The program is waiting to read data while a WAL segment is being created by copying an existing WAL segment.
	WALCopySync	The program is waiting for a WAL segment that is created by copying an existing WAL segment to reach stable storage.
	WALCopyWrite	The program is waiting to write data while a WAL segment is being created by copying an existing WAL segment.
	WALInitSync	The program is waiting for a newly initialized WAL file to reach stable storage.
	WALInitWrite	The program is waiting to write data while a new WAL file is being initialized.
	WALRead	The program is waiting to read data from a WAL file.
	WALSenderTimelineHistoryRead	The program is waiting to read data from a timeline history file while the walsender timeline command is being run.
WALSyncMethodAssign	The program is waiting for data to reach stable storage while the WAL sync method is being assigned.
WALWrite	The program is waiting to write data to a WAL file.

Note The polar_stat_activity view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

polar_stat_activity_rt

The polar_stat_activity_rt view provides real-time data of the status of all processes.


Parameter	Type	Description
`pid`	integer	The ID of the process.
`backend_type`	text	The type of the backend process.
`cpu_user`	bigint	The user CPU utilization of the backend server.
`cpu_sys`	bigint	The system CPU utilization of the backend server.
`rss`	bigint	The memory usage of the backend server.
`local_read_ps`	bigint	The cumulative sum of read operations on a local file system.
`local_write_ps`	bigint	The cumulative sum of write operations on a local file system.
`local_read_throughput`	bigint	The cumulative sum of the read throughput of a local file system.
`local_write_throughput`	bigint	The cumulative sum of the write throughput of a local file system.
`local_read_latency_ms`	double	The latency of the read operation on a local file system.
`local_write_latency_ms`	double	The latency of the write operation on a local file system.

Note The polar_stat_activity_rt view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

polar_delta
Important polar_delta is a function used to query the incremental values of a view. This function is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.
To use the polar_delta function, perform the following instructions:
1. Create a view that contains a column for dimensions and a column for values.
  - The name of the column for dimensions must start with d.
  - The name of the column for values must start with v.
2. Execute the following statements to query the view:
```
"select * from polar_delta (NULL::view_name)"
"\watch 1 select * from polar_delta (NULL::view_name)"
                            
```

Resources

CPU
Session-level CPU metrics are displayed in the polar_stat_activity view.
Shared memory
Other global data structures, such as the buffer pool and latch of PolarDB, are allocated to shared memory, which is applied at startup. You can obtain relevant information by querying the following view.
- polar_stat_shmem
  The polar_stat_shmem view provides detailed monitoring information about each type of shared memory. The following table describes the parameters in this view.
  Parameter Type Description
  shmname text The name of the shared memory.
  shmsize bigint The size of the shared memory. (Unit: bytes)
  shmtype text The type of the shared memory.
  Note The polar_stat_shmem view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.
- polar_stat_shmem_total_size
  The polar_stat_shmem_total_size view provides aggregated monitoring information about shared memory. The following table describes the parameters in this view.
  Parameter Type Description
  shmsize bigint The size of the shared memory. (Unit: bytes)
  shmtype text The type of the shared memory.
  Note The polar_stat_shmem_total_size view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

Parameter	Type	Description
`shmname`	text	The name of the shared memory.
`shmsize`	bigint	The size of the shared memory. (Unit: bytes)
`shmtype`	text	The type of the shared memory.

Parameter	Type	Description
`shmsize`	bigint	The size of the shared memory. (Unit: bytes)
`shmtype`	text	The type of the shared memory.

Session private memory

The private memory of PolarDB is dynamically allocated and released during the running process. The session-level memory metrics are displayed in the polar_stat_activity view. Context-dependent memory is the basic unit of dynamic memory management in PolarDB, which can be obtained by viewing the following functions and views.

polar_get_mcxt() function


Parameter	Type	Description
`pid`	integer	The ID of the session process.
`name`	text	The name of the context-dependent memory.
`level`	int	The level of the context-dependent memory.
`nblocks`	bigint	The number of the occupied blocks.
`freechunks`	bigint	The number of the free blocks.
`totalspace`	bigint	The total memory. (Unit: bytes)
`freespace`	bigint	The free memory. (Unit: bytes)

polar_backends_mcxt

The polar_backends_mcxt view provides information about context-dependent memory grouped by backend_type. The following table describes the parameters in this view.


Parameter	Type	Description
`pid`	integer	The ID of the process.
`name`	text	The name of the context-dependent memory.
`nblocks`	bigint	The number of the occupied blocks.
`freechunks`	bigint	The number of the free blocks.
`totalspace`	bigint	The total memory. (Unit: bytes)
`freespace`	bigint	The free memory. (Unit: bytes)

Note The polar_backends_mcxt view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

I/O

Session-level I/O metrics are displayed in the polar_stat_activity view. File-level I/O information and I/O latency distribution can be obtained by querying the following views.

polar_stat_io_info

The polar_stat_io_info view provides monitoring information about I/O requests grouped by file types. The following table describes the parameters in this view.


Parameter	Type	Description
`filetype`	text	The type of the file.
`fileloc`	text	The file system where the file resides. Currently, there are two types of file systems: local file system and shared file system Polar File System (PFS).
`open_count`	numeric	The number of times the file has been opened.
`open_latency_us`	double	The total latency for opening the file. (Unit: microseconds)
`close_count`	numeric	The number of times the file has been closed.
`read_count`	numeric	The number of times the file has been read.
`write_count`	numeric	The number of times the file has been written.
`read_throughput`	numeric	The read throughput of the file system.
`write_throughput`	numeric	The write throughput of the file system.
`read_latency_us`	double	The total latency for reading the file. (Unit: microseconds)
`write_latency_us`	double	The total latency for writing data to the file. (Unit: microseconds)
`seek_count`	numeric	The number of times that you called the seek() function.
`seek_latency_us`	double	The total latency for calling the seek() function.
`creat_count`	numeric	The number of times that you created a file.
`creat_latency_us`	double	The total latency for creating a file.
`fsync_count`	numeric	The number of times that you called the fsync() function.
`fsync_latency_us`	double	The total latency for calling the fsync() function.
`falloc_count`	numeric	The number of times that you called the fallocate() function.
`falloc_latency_us`	double	The total latency for calling the fallocate() function.

Note The polar_stat_io_info view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

polar_stat_io_latency

The polar_stat_io_latency view provides I/O latency information grouped by I/O types. The following table describes the parameters in this view.


Parameter	Type	Description
`iokind`	text	The type of the operation on a file. Valid values:`fsync`,`creat`,`seek`,`open`,`read`,`write`,`falloc`.
`num_lessthan200us`	numeric	The number of times that the I/O latency is less than 200 microseconds.
`num_lessthan400us`	numeric	The number of times that the I/O latency is greater than 200 microseconds but less than 400 microseconds.
`num_lessthan600us`	numeric	The number of times that the I/O latency is greater than 400 microseconds but less than 600 microseconds.
`num_lessthan800us`	numeric	The number of times that the I/O latency is greater than 600 microseconds but less than 800 microseconds.
`num_lessthan1ms`	numeric	The number of times that the I/O latency is greater than 800 microseconds but less than 1 millisecond.
`num_lessthan10ms`	numeric	The number of times that the I/O latency is greater than 1 millisecond but less than 10 milliseconds.
`num_lessthan100ms`	numeric	The number of times that the I/O latency is greater than 10 milliseconds but less than 100 milliseconds.
`num_morethan100ms`	numeric	The number of times that the I/O latency is greater than 100 milliseconds.

Note The polar_stat_io_latency view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

Network

You can query the following functions and views to obtain network monitoring information.

polar_proc_stat_network() function


Parameter	Type	Description
`pid`	bigint	The ID of the process.
`send_bytes`	bigint	The total number of bytes sent.
`send_count`	bigint	The cumulative number of sending times.
`recv_bytes`	bigint	The total number of bytes received.
`recv_count`	bigint	The cumulative number of receiving times.
`sendq`	bigint	The length of the sending queue over the socket.
`recvq`	bigint	The length of the receiving queue over the socket.
`cwnd`	bigint	The length of the sliding window of the socket.
`rtt`	bigint	The round-trip time (RTT) of TCP packets. (Unit: microseconds)
`retrans`	bigint	The cumulative number of times that data is retransmitted.
`tcpinfo_update_time`	bigint	The last update time of the TCP socket monitoring information. The value of this parameter is a second-level timestamp. It is updated every second. The aforementioned `sendq`, `recvq`, `cwnd`, `rtt`, and `retrans` metrics are updated regularly, and their update time is displayed by this parameter.

polar_stat_network

The polar_stat_network view provides network monitoring information. The following table describes the parameters in this view.


Parameter	Type	Description
`send_count`	bigint	The number of times that the system sends network traffic.
`send_bytes`	bigint	The amount of outbound traffic. (Unit: bytes)
`recv_count`	bigint	The number of times that the system receives network traffic.
`recv_bytes`	bigint	The amount of inbound traffic. (Unit: bytes)
`retrans`	bigint	The number of times that data is retransmitted.

Note The polar_stat_network view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

Lock

You can query the following views to obtain lock information.

LWLock

polar_stat_lwlock

The polar_stat_lwlock view provides monitoring information about locks. The following table describes the parameters in this view.


Parameter	Type	Description
`tranche`	smallint	The ID of the lock.
`name`	text	The name of the lock.
`sh_acquire_count`	bigint	The number of times that the system acquires a shared lock.
`ex_acquire_count`	bigint	The number of times that the system acquires an exclusive lock.
`block_count`	bigint	The number of blocks that occurred.
`lock_nums`	bigint	The number of lightweight locks.
`wait_time`	bigint	The total lock wait time.

Note The polar_stat_lwlock view is provided by the polar_monitor_preload plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor_preload statement.

Lock

pg_locks


Parameter	Type	Description
`locktype`	text	The type of the lockable object. Valid values: relation, extend, page, tuple, transactionid, virtualxid, object, userlock, and advisory.
`database`	oid	The OID of the database in which the object to be locked exists. If the object is a shared object, the value is 0. If the object is an XID, the value is NULL.
`relation`	oid	The OID of the relation to be locked. If the object to be locked is not a relation or is only part of a relation, the value is NULL.
`page`	integer	The page number of the page to be locked within the relation. If the object to be locked is not a tuple or a relation page, the value is NULL.
`tuple`	smallint	The tuple number of the tuple to be locked within the page. If the object to be locked is not a tuple, the value is NULL.
`virtualxid`	text	The virtual ID of the transaction to be locked. If the object to be locked is not a virtual XID, the value is NULL.
`transactionid`	xid	The ID of the transaction to be locked. If the object to be locked is not an XID, the value is NULL.
`classid`	oid	The OID of the system catalog that contains the object to be locked. If the object is not a general database object, the value is NULL.
`objid`	oid	The ID of the object to be locked within the system catalog. If the object is not a general database object, the value is NULL.
`objsubid`	smallint	The ID of the column to be locked. The values in classid and objid refer to the table itself. If the object to be locked is another general database object, the value is 0. If the object is not a general database object, the value is NULL.
`virtualtransaction`	text	The virtual ID of the transaction that holds the lock or is waiting for the lock.
`pid`	integer	The ID of the server process that holds the lock or is waiting for the lock. If this lock is held by a prepared transaction, the value is NULL.
`mode`	text	The name of the lock mode that is held or desired by this process.
`granted`	boolean	If the lock is held, the value is true. If the lock is waited for, the value is false.
`fastpath`	boolean	If the lock is obtained by using a fast path, the value is true. If the lock is obtained from the main lock table, the value is false.

polar_stat_lock

The polar_stat_lock view provides lock statistics. The following table describes the parameters in this view.


Parameter	Type	Description
`id`	integer	The primary key of the table.
`lock_type`	text	The type of the lock.
`invalid`	numeric	Indicates that the lock is invalid.
`accesssharelock`	numeric	The number of ACCESS SHARE locks.
`rowsharelock`	numeric	The number of ROW SHARE locks.
`rowexclusivelock`	numeric	The number of ROW EXCLUSIVE locks.
`shareupdateexclusivelock`	numeric	The number of SHARE UPDATE EXCLUSIVE locks.
`sharelock`	numeric	The number of SHARE locks.
`sharerowexclusivelock`	numeric	The number of SHARE ROW EXCLUSIVE locks.
`exclusivelock`	numeric	The number of EXCLUSIVE locks.
`accessexclusivelock`	numeric	The number of ACCESS EXCLUSIVE locks.
`block_count`	numeric	The number of blocks caused by locks.
`fastpath_count`	numeric	The number of fast-path locks on a local system.
`wait_time`	numeric	The lock wait time.

Note The polar_stat_lock view is provided by the polar_monitor_preload plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor_preload statement.

SLRU

You can query the polar_stat_slru() view to obtain monitoring information about Segmented Least Recently Used (SLRU) cache data.

polar_stat_slru()

The polar_stat_slru() view provides monitoring information about SLRU cache data. The following table describes the parameters in this view.


Parameter	Type	Description
`slru_type`	text	The primary key of the table.
`slots_number`	integer	The number of pages.
`valid_pages`	integer	The number of the used pages.
`empty_pages`	integer	The number of the empty pages.
`reading_pages`	integer	The number of the pages on which the read operations are being performed.
`writing_pages`	integer	The number of the pages on which the write operations are being performed.
`wait_readings`	integer	The number of the wait events for read operations.
`wait_writings`	integer	The number of the wait events for write operations.
`read_count`	bigint	The number of the read operations.
`read_only_count`	bigint	The number of the read-only operations.
`read_upgrade_count`	bigint	The number of the read operations in the upgrade mode.
`victim_count`	bigint	The number of times that data is evicted from caches.
`victim_write_count`	bigint	The number of times that data is evicted from caches and then is written to disks.
`write_count`	bigint	The number of the write operations.
`zero_count`	bigint	The number of times that the cache data is cleared.
`flush_count`	bigint	The number of the flush operations.
`truncate_count`	bigint	The number of the truncate operations.
`storage_read_count`	bigint	The number of times that you read data from disks.
`storage_write_count`	bigint	The number of times that you write data to disks.

Note The polar_stat_slru view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

cgroup
You can query the following views to obtain information about system resources.
- polar_stat_cgroup
  The polar_stat_cgroup provides statistics about system resources and control groups (cgroups). The following table describes the parameters in this view.
  Parameter Type Description
  subtype text The type of the cgroup. Valid values: IO, Memory, and CPU.
  infotype text The information about the cgroup.
  count bigint The statistics about the used resources.
  Note The polar_stat_cgroup view is provided by the polar_monitor_preload and polar_monitor plug-ins. These plug-ins are built into PolarDB for PostgreSQL. To activate the plug-ins, execute the create extension polar_monitor_preload and create extension polar_monitor statements.
- polar_cgroup_quota
  The polar_cgroup_quota provides statistics about system resources and cgroup quotas. The following table describes the parameters in this view.
  Parameter Type Description
  subtype text The type of a cgroup. Valid values: IO, Memory, and CPU.
  infotype text The information about cgroups.
  count bigint The quota of a cgroup.
  Note The polar_cgroup_quota view is provided by the polar_monitor plug-in. This plug-in is built into PolarDB for PostgreSQL. To activate this plug-in, execute the create extension polar_monitor statement.

Parameter	Type	Description
`subtype`	text	The type of the cgroup. Valid values: IO, Memory, and CPU.
`infotype`	text	The information about the cgroup.
`count`	bigint	The statistics about the used resources.

Parameter	Type	Description
`subtype`	text	The type of a cgroup. Valid values: IO, Memory, and CPU.
`infotype`	text	The information about cgroups.
`count`	bigint	The quota of a cgroup.