Performance Insights counter metrics - Amazon Aurora

Performance Insights counter metrics

Counter metrics are operating system and database performance metrics in the Performance Insights dashboard. To help identify and analyze performance problems, you can correlate counter metrics with DB load.

Performance Insights operating system counters

The following operating system counters, which are prefixed with os, are available with Performance Insights for Aurora PostgreSQL. You can find definitions for these metrics in Viewing OS metrics using CloudWatch Logs.

Counter Type Metric
active memory
buffers memory os.memory.buffers
cached memory os.memory.cached
dirty memory os.memory.dirty
free memory
hugePagesFree memory os.memory.hugePagesFree
hugePagesRsvd memory os.memory.hugePagesRsvd
hugePagesSize memory os.memory.hugePagesSize
hugePagesSurp memory os.memory.hugePagesSurp
hugePagesTotal memory os.memory.hugePagesTotal
inactive memory os.memory.inactive
mapped memory os.memory.mapped
pageTables memory os.memory.pageTables
slab memory os.memory.slab
total memory
writeback memory os.memory.writeback
guest cpuUtilization os.cpuUtilization.guest
idle cpuUtilization os.cpuUtilization.idle
irq cpuUtilization os.cpuUtilization.irq
nice cpuUtilization os.cpuUtilization.nice
steal cpuUtilization os.cpuUtilization.steal
system cpuUtilization os.cpuUtilization.system
total cpuUtilization
user cpuUtilization os.cpuUtilization.user
wait cpuUtilization os.cpuUtilization.wait
avgQueueLen diskIO os.diskIO.avgQueueLen
avgReqSz diskIO os.diskIO.avgReqSz
await diskIO os.diskIO.await
readIOsPS diskIO os.diskIO.readIOsPS
readKb diskIO os.diskIO.readKb
readKbPS diskIO os.diskIO.readKbPS
rrqmPS diskIO os.diskIO.rrqmPS
tps diskIO os.diskIO.tps
util diskIO os.diskIO.util
writeIOsPS diskIO os.diskIO.writeIOsPS
writeKb diskIO os.diskIO.writeKb
writeKbPS diskIO os.diskIO.writeKbPS
wrqmPS diskIO os.diskIO.wrqmPS
blocked tasks os.tasks.blocked
running tasks os.tasks.running
sleeping tasks os.tasks.sleeping
stopped tasks os.tasks.stopped
total tasks
zombie tasks os.tasks.zombie
one loadAverageMinute
fifteen loadAverageMinute os.loadAverageMinute.fifteen
five loadAverageMinute os.loadAverageMinute.five
cached swap os.swap.cached
free swap
in swap
out swap os.swap.out
total swap
maxFiles fileSys os.fileSys.maxFiles
usedFiles fileSys os.fileSys.usedFiles
usedFilePercent fileSys os.fileSys.usedFilePercent
usedPercent fileSys os.fileSys.usedPercent
used fileSys os.fileSys.used
total fileSys
rx network
tx network
acuUtilization general os.general.acuUtilization
maxConfiguredAcu general os.general.maxConfiguredAcu
minConfiguredAcu general os.general.minConfiguredAcu
numVCPUs general os.general.numVCPUs
serverlessDatabaseCapacity general os.general.serverlessDatabaseCapacity

Performance Insights counters for Aurora MySQL

The following database counters are available with Performance Insights for Aurora MySQL.

Native counters for Aurora MySQL

Native metrics are defined by the database engine and not by Amazon Aurora. You can find definitions for these native metrics in Server status variables in the MySQL documentation.

Counter Type Unit Metric
Com_analyze SQL Queries per second db.SQL.Com_analyze
Com_optimize SQL Queries per second db.SQL.Com_optimize
Com_select SQL Queries per second db.SQL.Com_select
Innodb_rows_deleted SQL Rows per second db.SQL.Innodb_rows_deleted
Innodb_rows_inserted SQL Rows per second db.SQL.Innodb_rows_inserted
Innodb_rows_read SQL Rows per second db.SQL.Innodb_rows_read
Innodb_rows_updated SQL Rows per second db.SQL.Innodb_rows_updated
Queries SQL Queries per second db.SQL.Queries
Questions SQL Queries per second db.SQL.Questions
Select_full_join SQL Queries per second db.SQL.Select_full_join
Select_full_range_join SQL Queries per second db.SQL.Select_full_range_join
Select_range SQL Queries per second db.SQL.Select_range
Select_range_check SQL Queries per second db.SQL.Select_range_check
Select_scan SQL Queries per second db.SQL.Select_scan
Slow_queries SQL Queries per second db.SQL.Slow_queries
Sort_merge_passes SQL Queries per second db.SQL.Sort_merge_passes
Sort_range SQL Queries per second db.SQL.Sort_range
Sort_rows SQL Queries per second db.SQL.Sort_rows
Sort_scan SQL Queries per second db.SQL.Sort_scan
Total_query_time SQL Milliseconds db.SQL.Total_query_time
Table_locks_immediate Locks Requests per second db.Locks.Table_locks_immediate
Table_locks_waited Locks Requests per second db.Locks.Table_locks_waited
Innodb_row_lock_time Locks Milliseconds (average) db.Locks.Innodb_row_lock_time
Aborted_clients Users Connections db.Users.Aborted_clients
Aborted_connects Users Connections db.Users.Aborted_connects
Connections Users Connections db.Users.Connections
External_threads_connected Users Connections db.Users.External_threads_connected
Threads_connected Users Connections db.Users.Threads_connected
Threads_created Users Connections db.Users.Threads_created
Threads_running Users Connections db.Users.Threads_running
Created_tmp_disk_tables Temp Tables per second db.Temp.Created_tmp_disk_tables
Created_tmp_tables Temp Tables per second db.Temp.Created_tmp_tables
Innodb_buffer_pool_pages_data Cache Pages db.Cache.Innodb_buffer_pool_pages_data
Innodb_buffer_pool_pages_total Cache Pages db.Cache.Innodb_buffer_pool_pages_total
Innodb_buffer_pool_read_requests Cache Pages per second db.Cache.Innodb_buffer_pool_read_requests
Innodb_buffer_pool_reads Cache Pages per second db.Cache.Innodb_buffer_pool_reads
Opened_tables Cache Tables db.Cache.Opened_tables
Opened_table_definitions Cache Tables db.Cache.Opened_table_definitions
Qcache_hits Cache Queries db.Cache.Qcache_hits

Non-native counters for Aurora MySQL

Non-native counter metrics are counters defined by Amazon RDS. A non-native metric can be a metric that you get with a specific query. A non-native metric also can be a derived metric, where two or more native counters are used in calculations for ratios, hit rates, or latencies.

Counter Type Metric Description Definition
innodb_buffer_pool_hits Cache db.Cache.innoDB_buffer_pool_hits The number of reads that InnoDB could satisfy from the buffer pool. innodb_buffer_pool_read_requests - innodb_buffer_pool_reads
innodb_buffer_pool_hit_rate Cache db.Cache.innoDB_buffer_pool_hit_rate The percentage of reads that InnoDB could satisfy from the buffer pool. 100 * innodb_buffer_pool_read_requests / (innodb_buffer_pool_read_requests + innodb_buffer_pool_reads)
innodb_buffer_pool_usage Cache db.Cache.innoDB_buffer_pool_usage

The percentage of the InnoDB buffer pool that contains data (pages).


When using compressed tables, this value can vary. For more information, see the information about Innodb_buffer_pool_pages_data and Innodb_buffer_pool_pages_total in Server status sariables in the MySQL documentation.

Innodb_buffer_pool_pages_data / Innodb_buffer_pool_pages_total * 100.0
query_cache_hit_rate Cache db.Cache.query_cache_hit_rate The hit ratio for the MySQL result set cache (query cache). Qcache_hits / (QCache_hits + Com_select) * 100
innodb_rows_changed SQL db.SQL.innodb_rows_changed The total InnoDB row operations. db.SQL.Innodb_rows_inserted + db.SQL.Innodb_rows_deleted + db.SQL.Innodb_rows_updated
active_transactions Transactions db.Transactions.active_transactions The total active transactions. SELECT COUNT(1) AS active_transactions FROM INFORMATION_SCHEMA.INNODB_TRX
trx_rseg_history_len Transactions db.Transactions.trx_rseg_history_len A list of the undo log pages for committed transactions that is maintained by the InnoDB transaction system to implement multi-version concurrency control. For more information about undo log records details, see in the MySQL documentation. SELECT COUNT AS trx_rseg_history_len FROM INFORMATION_SCHEMA.INNODB_METRICS WHERE NAME='trx_rseg_history_len'
innodb_deadlocks Locks db.Locks.innodb_deadlocks The total number of deadlocks. SELECT COUNT AS innodb_deadlocks FROM INFORMATION_SCHEMA.INNODB_METRICS WHERE NAME='lock_deadlocks'
innodb_lock_timeouts Locks db.Locks.innodb_lock_timeouts The total number of deadlocks that timed out. SELECT COUNT AS innodb_lock_timeouts FROM INFORMATION_SCHEMA.INNODB_METRICS WHERE NAME='lock_timeouts'
innodb_row_lock_waits Locks db.Locks.innodb_row_lock_waits The total number of row locks that resulted in a wait. SELECT COUNT AS innodb_row_lock_waits FROM INFORMATION_SCHEMA.INNODB_METRICS WHERE NAME='lock_row_lock_waits'

Performance Insights counters for Aurora PostgreSQL

The following database counters are available with Performance Insights for Aurora PostgreSQL.

Native Counters for Aurora PostgreSQL

Native metrics are defined by the database engine and not by Amazon Aurora. You can find definitions for these native metrics in Viewing Statistics in the PostgreSQL documentation.

Counter Type Unit Metric
queries_started SQL Queries per second db.SQL.queries
total_query_time SQL Milliseconds db.SQL.total_query_time
tup_deleted SQL Tuples per second db.SQL.tup_deleted
tup_fetched SQL Tuples per second db.SQL.tup_fetched
tup_inserted SQL Tuples per second db.SQL.tup_inserted
tup_returned SQL Tuples per second db.SQL.tup_returned
tup_updated SQL Tuples per second db.SQL.tup_updated
blks_hit Cache Blocks per second db.Cache.blks_hit
buffers_alloc Cache Blocks per second db.Cache.buffers_alloc
buffers_checkpoint Checkpoint Blocks per second db.Checkpoint.buffers_checkpoint
checkpoints_req Checkpoint Checkpoints per minute db.Checkpoint.checkpoints_req
checkpoint_sync_time Checkpoint Milliseconds per checkpoint db.Checkpoint.checkpoint_sync_time
checkpoints_timed Checkpoint Checkpoints per minute db.Checkpoint.checkpoints_timed
checkpoint_write_time Checkpoint Milliseconds per checkpoint db.Checkpoint.checkpoint_write_time
maxwritten_clean Checkpoint Bgwriter clean stops per minute db.Checkpoint.maxwritten_clean
deadlocks Concurrency Deadlocks per minute db.Concurrency.deadlocks
blk_read_time I/O Milliseconds db.IO.blk_read_time
blks_read I/O Blocks per second db.IO.blks_read
buffers_backend I/O Blocks per second db.IO.buffers_backend
buffers_backend_fsync I/O Blocks per second db.IO.buffers_backend_fsync
buffers_clean I/O Blocks per second db.IO.buffers_clean
idle_in_transaction_aborted_count State Sessions db.State.idle_in_transaction_aborted_count
idle_in_transaction_count State Sessions db.State.idle_in_transaction_count
idle_in_transaction_max_time State Seconds db.State.idle_in_transaction_max_time
temp_bytes Temp Bytes per second db.Temp.temp_bytes
temp_files Temp Files per minute db.Temp.temp_files
active_transactions Transactions Transactions db.Transactions.active_transactions
blocked_transactions Transactions Transactions db.Transactions.blocked_transactions
duration_commits Transactions Milliseconds db.Transactions.duration_commits
max_used_xact_ids Transactions Transactions db.Transactions.max_used_xact_ids
xact_commit Transactions Commits per second db.Transactions.xact_commit
xaxt_rollback Transactions Rollbacks per second db.Transactions.xact_rollback
numbackends User Connections db.User.numbackends
total_auth_attempts User Connections per minute db.User.total_auth_attempts
archived_count WAL Files per minute db.WAL.archived_count
archive_failed_count WAL Files per minute db.WAL.archive_failed_count

Non-native counters for Aurora PostgreSQL

Non-native counter metrics are counters defined by Amazon Aurora. A non-native metric can be a metric that you get with a specific query. A non-native metric also can be a derived metric, where two or more native counters are used in calculations for ratios, hit rates, or latencies.

Counter Type Metric Description Definition
logical_reads SQL db.SQL.logical_reads The total number of blocks hit and read. blks_hit + blks_read
checkpoint_sync_latency Checkpoint db.Checkpoint.checkpoint_sync_latency The total amount of time that has been spent in the portion of checkpoint processing where files are synchronized to disk. checkpoint_sync_time / (checkpoints_timed + checkpoints_req)
checkpoint_write_latency Checkpoint db.Checkpoint.checkpoint_write_latency The total amount of time that has been spent in the portion of checkpoint processing where files are written to disk. checkpoint_write_time / (checkpoints_timed + checkpoints_req)
read_latency I/O db.IO.read_latency The time spent reading data file blocks by backends in this instance. blk_read_time / blks_read
commit_latency Transactions db.Transactions.commit_latency The average duration of commit operations. db.Transactions.duration_commits / db.Transactions.xact_commit