Custom Chart Debug Page

On this page Carat arrow pointing down

The Custom Chart debug page in the DB Console lets you create one or multiple custom charts showing any combination of available metrics.

The definition of the customized dashboard is encoded in the URL. To share the dashboard with someone, send them the URL. Like any other URL, it can be bookmarked, sit in a pinned tab in your browser, etc.

To view the Custom Chart page, access the DB Console, click Advanced Debug In the left-hand navigation bar, and in the Reports section select Custom Time Series Chart.

Use the Custom Chart page

DB Console

On the Custom Chart page, you can set the time span for all charts, add new custom charts, and customize each chart:

  • To set the time span for the page, use the dropdown menu above the charts and select the desired time span. In addition, once you have selected a metric to display, you can drag within the chart itself to a set a custom time range.

  • To add a chart, click Add Chart and customize the new chart.

  • To customize each chart, use the Units dropdown menu to set the units to display. Then use the table below the chart to select the metrics being queried, and how they'll be combined and displayed. Options include:

    Column Description
    Metric Name How the system refers to this metric, e.g., sql.bytesin.
    Downsampler

    The "Downsampler" operation is used to combine the individual datapoints over the longer period into a single datapoint. We store one data point every ten seconds, but for queries over long time spans the backend lowers the resolution of the returned data, perhaps only returning one data point for every minute, five minutes, or even an entire hour in the case of the 30 day view.

    Options:

    • AVG: Returns the average value over the time period.
    • MIN: Returns the lowest value seen.
    • MAX: Returns the highest value seen.
    • SUM: Returns the sum of all values seen.

    Aggregator

    Used to combine data points from different nodes. It has the same operations available as the Downsampler.

    Options:

    • AVG: Returns the average value over the time period.
    • MIN: Returns the lowest value seen.
    • MAX: Returns the highest value seen.
    • SUM: Returns the sum of all values seen.

    Rate

    Determines how to display the rate of change during the selected time period.

    Options:

    • Normal: Returns the actual recorded value.
    • Rate: Returns the rate of change of the value per second.
    • Non-negative Rate: Returns the rate-of-change, but returns 0 instead of negative values. A large number of the stats we track are actually tracked as monotonically increasing counters so each sample is just the total value of that counter. The rate of change of that counter represents the rate of events being counted, which is usually what you want to graph. "Non-negative Rate" is needed because the counters are stored in memory, and thus if a node resets it goes back to zero (whereas normally they only increase).

    Source The set of nodes being queried, which is either:
    • The entire cluster.
    • A single, named node.
    Per Node If checked, the chart will show a line for each node's value of this metric.

Examples

Query user and system CPU usage

DB Console

To compare system vs. userspace CPU usage, select the following values under Metric Name:

  • sys.cpu.sys.percent
  • sys.cpu.user.percent

The Y-axis label is the Count. A count of 1 represents 100% utilization. The Aggregator of Sum can show the count to be above 1, which would mean CPU utilization is greater than 100%.

Checking Per Node displays statistics for each node, which could show whether an individual node's CPU usage was higher or lower than the average.

Essential Metrics to Monitor

For important metrics to visualize in a custom dashboard, refer to:

Available metrics

Note:

Some of the metrics listed below are already visible in other areas of the DB Console.

CockroachDB Metric Name Description Type Unit
addsstable.delay.total
Amount by which evaluation of AddSSTable requests was delayed COUNTER NANOSECONDS
addsstable.proposals
Number of SSTable ingestions proposed (i.e. sent to Raft by lease holders) COUNTER COUNT
admission.io.overload
1-normalized float indicating whether IO admission control considers the store as overloaded with respect to compaction out of L0 (considers sub-level and file counts). GAUGE PERCENT
admission.wait_durations.kv
Wait time durations for requests that waited HISTOGRAM NANOSECONDS
admission.wait_durations.kv-stores
Wait time durations for requests that waited HISTOGRAM NANOSECONDS
auth.cert.conn.latency
Latency to establish and authenticate a SQL connection using certificate HISTOGRAM NANOSECONDS
auth.gss.conn.latency
Latency to establish and authenticate a SQL connection using GSS HISTOGRAM NANOSECONDS
auth.jwt.conn.latency
Latency to establish and authenticate a SQL connection using JWT Token HISTOGRAM NANOSECONDS
auth.ldap.conn.latency
Latency to establish and authenticate a SQL connection using LDAP HISTOGRAM NANOSECONDS
auth.ldap.conn.latency.internal
Internal Auth Latency to establish and authenticate a SQL connection using LDAP(excludes external LDAP calls) HISTOGRAM NANOSECONDS
auth.password.conn.latency
Latency to establish and authenticate a SQL connection using password HISTOGRAM NANOSECONDS
auth.scram.conn.latency
Latency to establish and authenticate a SQL connection using SCRAM HISTOGRAM NANOSECONDS
capacity
Total storage capacity GAUGE BYTES
capacity.available
Available storage capacity GAUGE BYTES
capacity.used
Used storage capacity GAUGE BYTES
changefeed.backfill_count
Number of changefeeds currently executing backfill GAUGE COUNT
changefeed.commit_latency
Event commit latency: a difference between event MVCC timestamp and the time it was acknowledged by the downstream sink. If the sink batches events, then the difference between the oldest event in the batch and acknowledgement is recorded. Excludes latency during backfill. HISTOGRAM NANOSECONDS
changefeed.emitted_bytes
Bytes emitted by all feeds COUNTER BYTES
changefeed.emitted_messages
Messages emitted by all feeds COUNTER COUNT
changefeed.error_retries
Total retryable errors encountered by all changefeeds COUNTER COUNT
changefeed.failures
Total number of changefeed jobs which have failed COUNTER COUNT
changefeed.max_behind_nanos
The most any changefeed's persisted checkpoint is behind the present GAUGE NANOSECONDS
changefeed.running
Number of currently running changefeeds, including sinkless GAUGE COUNT
clock-offset.meannanos
Mean clock offset with other nodes GAUGE NANOSECONDS
distsender.errors.notleaseholder
Number of NotLeaseHolderErrors encountered from replica-addressed RPCs COUNTER COUNT
distsender.rpc.sent.nextreplicaerror
Number of replica-addressed RPCs sent due to per-replica errors COUNTER COUNT
exec.latency
Latency of batch KV requests (including errors) executed on this node.

This measures requests already addressed to a single replica, from the moment at which they arrive at the internal gRPC endpoint to the moment at which the response (or an error) is returned.

This latency includes in particular commit waits, conflict resolution and replication, and end-users can easily produce high measurements via long-running transactions that conflict with foreground traffic. This metric thus does not provide a good signal for understanding the health of the KV layer.

HISTOGRAM NANOSECONDS
go.scheduler_latency
Go scheduling latency HISTOGRAM NANOSECONDS
intentcount
Count of intent keys GAUGE COUNT
jobs.auto_create_partial_stats.currently_paused
Number of auto_create_partial_stats jobs currently considered Paused GAUGE COUNT
jobs.auto_create_partial_stats.currently_running
Number of auto_create_partial_stats jobs currently running in Resume or OnFailOrCancel state GAUGE COUNT
jobs.auto_create_partial_stats.resume_failed
Number of auto_create_partial_stats jobs which failed with a non-retriable error COUNTER COUNT
jobs.auto_create_stats.currently_paused
Number of auto_create_stats jobs currently considered Paused GAUGE COUNT
jobs.auto_create_stats.currently_running
Number of auto_create_stats jobs currently running in Resume or OnFailOrCancel state GAUGE COUNT
jobs.auto_create_stats.resume_failed
Number of auto_create_stats jobs which failed with a non-retriable error COUNTER COUNT
jobs.backup.currently_paused
Number of backup jobs currently considered Paused GAUGE COUNT
jobs.backup.currently_running
Number of backup jobs currently running in Resume or OnFailOrCancel state GAUGE COUNT
jobs.changefeed.currently_paused
Number of changefeed jobs currently considered Paused GAUGE COUNT
jobs.changefeed.protected_age_sec
The age of the oldest PTS record protected by changefeed jobs GAUGE SECONDS
jobs.create_stats.currently_running
Number of create_stats jobs currently running in Resume or OnFailOrCancel state GAUGE COUNT
jobs.row_level_ttl.currently_paused
Number of row_level_ttl jobs currently considered Paused GAUGE COUNT
jobs.row_level_ttl.currently_running
Number of row_level_ttl jobs currently running in Resume or OnFailOrCancel state GAUGE COUNT
jobs.row_level_ttl.delete_duration
Duration for delete requests during row level TTL. HISTOGRAM NANOSECONDS
jobs.row_level_ttl.num_active_spans
Number of active spans the TTL job is deleting from. GAUGE COUNT
jobs.row_level_ttl.resume_completed
Number of row_level_ttl jobs which successfully resumed to completion COUNTER COUNT
jobs.row_level_ttl.resume_failed
Number of row_level_ttl jobs which failed with a non-retriable error COUNTER COUNT
jobs.row_level_ttl.rows_deleted
Number of rows deleted by the row level TTL job. COUNTER COUNT
jobs.row_level_ttl.rows_selected
Number of rows selected for deletion by the row level TTL job. COUNTER COUNT
jobs.row_level_ttl.select_duration
Duration for select requests during row level TTL. HISTOGRAM NANOSECONDS
jobs.row_level_ttl.span_total_duration
Duration for processing a span during row level TTL. HISTOGRAM NANOSECONDS
jobs.row_level_ttl.total_expired_rows
Approximate number of rows that have expired the TTL on the TTL table. GAUGE COUNT
jobs.row_level_ttl.total_rows
Approximate number of rows on the TTL table. GAUGE COUNT
kv.concurrency.locks
Number of active locks held in lock tables. Does not include replicated locks (intents) that are not held in memory GAUGE COUNT
kv.rangefeed.catchup_scan_nanos
Time spent in RangeFeed catchup scan COUNTER NANOSECONDS
leases.epoch
Number of replica leaseholders using epoch-based leases GAUGE COUNT
leases.expiration
Number of replica leaseholders using expiration-based leases GAUGE COUNT
leases.leader
Number of replica leaseholders using leader leases GAUGE COUNT
leases.liveness
Number of replica leaseholders for the liveness range(s) GAUGE COUNT
leases.transfers.error
Number of failed lease transfers COUNTER COUNT
leases.transfers.success
Number of successful lease transfers COUNTER COUNT
livebytes
Number of bytes of live data (keys plus values) GAUGE BYTES
liveness.heartbeatfailures
Number of failed node liveness heartbeats from this node COUNTER COUNT
liveness.heartbeatlatency
Node liveness heartbeat latency HISTOGRAM NANOSECONDS
liveness.livenodes
Number of live nodes in the cluster (will be 0 if this node is not itself live) GAUGE COUNT
logical_replication.commit_latency
Event commit latency: a difference between event MVCC timestamp and the time it was flushed into disk. If we batch events, then the difference between the oldest event in the batch and flush is recorded HISTOGRAM NANOSECONDS
logical_replication.events_dlqed
Row update events sent to DLQ COUNTER COUNT
logical_replication.events_ingested
Events ingested by all replication jobs COUNTER COUNT
logical_replication.logical_bytes
Logical bytes (sum of keys + values) received by all replication jobs COUNTER BYTES
logical_replication.replicated_time_seconds
The replicated time of the logical replication stream in seconds since the unix epoch. GAUGE SECONDS
physical_replication.logical_bytes
Logical bytes (sum of keys + values) ingested by all replication jobs COUNTER BYTES
physical_replication.replicated_time_seconds
The replicated time of the physical replication stream in seconds since the unix epoch. GAUGE SECONDS
queue.gc.pending
Number of pending replicas in the MVCC GC queue GAUGE COUNT
queue.gc.process.failure
Number of replicas which failed processing in the MVCC GC queue COUNTER COUNT
queue.lease.pending
Number of pending replicas in the replica lease queue GAUGE COUNT
queue.merge.pending
Number of pending replicas in the merge queue GAUGE COUNT
queue.merge.process.failure
Number of replicas which failed processing in the merge queue COUNTER COUNT
queue.merge.process.success
Number of replicas successfully processed by the merge queue COUNTER COUNT
queue.merge.processingnanos
Nanoseconds spent processing replicas in the merge queue COUNTER NANOSECONDS
queue.raftlog.pending
Number of pending replicas in the Raft log queue GAUGE COUNT
queue.raftlog.process.failure
Number of replicas which failed processing in the Raft log queue COUNTER COUNT
queue.raftlog.process.success
Number of replicas successfully processed by the Raft log queue COUNTER COUNT
queue.raftlog.processingnanos
Nanoseconds spent processing replicas in the Raft log queue COUNTER NANOSECONDS
queue.replicagc.pending
Number of pending replicas in the replica GC queue GAUGE COUNT
queue.replicagc.process.failure
Number of replicas which failed processing in the replica GC queue COUNTER COUNT
queue.replicagc.process.success
Number of replicas successfully processed by the replica GC queue COUNTER COUNT
queue.replicate.pending
Number of pending replicas in the replicate queue GAUGE COUNT
queue.replicate.process.failure
Number of replicas which failed processing in the replicate queue COUNTER COUNT
queue.replicate.process.success
Number of replicas successfully processed by the replicate queue COUNTER COUNT
queue.replicate.replacedecommissioningreplica.error
Number of failed decommissioning replica replacements processed by the replicate queue COUNTER COUNT
raft.scheduler.latency
Queueing durations for ranges waiting to be processed by the Raft scheduler.

This histogram measures the delay from when a range is registered with the scheduler for processing to when it is actually processed. This does not include the duration of processing.

HISTOGRAM NANOSECONDS
raftlog.behind
Number of Raft log entries followers on other stores are behind.

This gauge provides a view of the aggregate number of log entries the Raft leaders on this node think the followers are behind. Since a raft leader may not always have a good estimate for this information for all of its followers, and since followers are expected to be behind (when they are not required as part of a quorum) and the aggregate thus scales like the count of such followers, it is difficult to meaningfully interpret this metric.

GAUGE COUNT
range.adds
Number of range additions COUNTER COUNT
range.merges
Number of range merges COUNTER COUNT
range.snapshots.send-queue
Number of snapshots queued to send GAUGE COUNT
range.splits
Number of range splits COUNTER COUNT
ranges
Number of ranges GAUGE COUNT
ranges.decommissioning
Number of ranges with at lease one replica on a decommissioning node GAUGE COUNT
ranges.unavailable
Number of ranges with fewer live replicas than needed for quorum GAUGE COUNT
ranges.underreplicated
Number of ranges with fewer live replicas than the replication target GAUGE COUNT
rebalancing.cpunanospersecond
Average CPU nanoseconds spent on processing replica operations in the last 30 minutes. GAUGE NANOSECONDS
rebalancing.lease.transfers
Number of lease transfers motivated by store-level load imbalances COUNTER COUNT
rebalancing.queriespersecond
Number of kv-level requests received per second by the store, considering the last 30 minutes, as used in rebalancing decisions. GAUGE COUNT
rebalancing.range.rebalances
Number of range rebalance operations motivated by store-level load imbalances COUNTER COUNT
rebalancing.replicas.cpunanospersecond
Histogram of average CPU nanoseconds spent on processing replica operations in the last 30 minutes. HISTOGRAM NANOSECONDS
rebalancing.replicas.queriespersecond
Histogram of average kv-level requests received per second by replicas on the store in the last 30 minutes. HISTOGRAM COUNT
replicas
Number of replicas GAUGE COUNT
replicas.leaseholders
Number of lease holders GAUGE COUNT
requests.slow.latch
Number of requests that have been stuck for a long time acquiring latches.

Latches moderate access to the KV keyspace for the purpose of evaluating and replicating commands. A slow latch acquisition attempt is often caused by another request holding and not releasing its latches in a timely manner. This in turn can either be caused by a long delay in evaluation (for example, under severe system overload) or by delays at the replication layer.

This gauge registering a nonzero value usually indicates a serious problem and should be investigated.

GAUGE COUNT
requests.slow.lease
Number of requests that have been stuck for a long time acquiring a lease.

This gauge registering a nonzero value usually indicates range or replica unavailability, and should be investigated. In the common case, we also expect to see 'requests.slow.raft' to register a nonzero value, indicating that the lease requests are not getting a timely response from the replication layer.

GAUGE COUNT
requests.slow.raft
Number of requests that have been stuck for a long time in the replication layer.

An (evaluated) request has to pass through the replication layer, notably the quota pool and raft. If it fails to do so within a highly permissive duration, the gauge is incremented (and decremented again once the request is either applied or returns an error).

A nonzero value indicates range or replica unavailability, and should be investigated.

GAUGE COUNT
rocksdb.block.cache.hits
Count of block cache hits COUNTER COUNT
rocksdb.block.cache.misses
Count of block cache misses COUNTER COUNT
rocksdb.compactions
Number of table compactions COUNTER COUNT
rocksdb.read-amplification
Number of disk reads per query GAUGE CONST
round-trip-latency
Distribution of round-trip latencies with other nodes.

This only reflects successful heartbeats and measures gRPC overhead as well as possible head-of-line blocking. Elevated values in this metric may hint at network issues and/or saturation, but they are no proof of them. CPU overload can similarly elevate this metric. The operator should look towards OS-level metrics such as packet loss, retransmits, etc, to conclusively diagnose network issues. Heartbeats are not very frequent (~seconds), so they may not capture rare or short-lived degradations.

HISTOGRAM NANOSECONDS
rpc.connection.avg_round_trip_latency
Sum of exponentially weighted moving average of round-trip latencies, as measured through a gRPC RPC.

Since this metric is based on gRPC RPCs, it is affected by application-level processing delays and CPU overload effects. See rpc.connection.tcp_rtt for a metric that is obtained from the kernel's TCP stack.

Dividing this Gauge by rpc.connection.healthy gives an approximation of average latency, but the top-level round-trip-latency histogram is more useful. Instead, users should consult the label families of this metric if they are available (which requires prometheus and the cluster setting 'server.child_metrics.enabled'); these provide per-peer moving averages.

This metric does not track failed connection. A failed connection's contribution is reset to zero.

GAUGE NANOSECONDS
rpc.connection.failures
Counter of failed connections.

This includes both the event in which a healthy connection terminates as well as unsuccessful reconnection attempts.

Connections that are terminated as part of local node shutdown are excluded. Decommissioned peers are excluded.

COUNTER COUNT
rpc.connection.healthy
Gauge of current connections in a healthy state (i.e. bidirectionally connected and heartbeating) GAUGE COUNT
rpc.connection.healthy_nanos
Gauge of nanoseconds of healthy connection time

On the prometheus endpoint scraped with the cluster setting 'server.child_metrics.enabled' set, the constituent parts of this metric are available on a per-peer basis and one can read off for how long a given peer has been connected

GAUGE NANOSECONDS
rpc.connection.heartbeats
Counter of successful heartbeats. COUNTER COUNT
rpc.connection.tcp_rtt
Kernel-level TCP round-trip time as measured by the Linux TCP stack.

This metric reports the smoothed round-trip time (SRTT) as maintained by the kernel's TCP implementation. Unlike application-level RPC latency measurements, this reflects pure network latency and is less affected by CPU overload effects.

This metric is only available on Linux.

GAUGE NANOSECONDS
rpc.connection.tcp_rtt_var
Kernel-level TCP round-trip time variance as measured by the Linux TCP stack.

This metric reports the smoothed round-trip time variance (RTTVAR) as maintained by the kernel's TCP implementation. This measures the stability of the connection latency.

This metric is only available on Linux.

GAUGE NANOSECONDS
rpc.connection.unhealthy
Gauge of current connections in an unhealthy state (not bidirectionally connected or heartbeating) GAUGE COUNT
rpc.connection.unhealthy_nanos
Gauge of nanoseconds of unhealthy connection time.

On the prometheus endpoint scraped with the cluster setting 'server.child_metrics.enabled' set, the constituent parts of this metric are available on a per-peer basis and one can read off for how long a given peer has been unreachable

GAUGE NANOSECONDS
schedules.BACKUP.failed
Number of BACKUP jobs failed COUNTER COUNT
schedules.BACKUP.last-completed-time
The unix timestamp of the most recently completed backup by a schedule specified as maintaining this metric GAUGE TIMESTAMP_SEC
schedules.scheduled-row-level-ttl-executor.failed
Number of scheduled-row-level-ttl-executor jobs failed COUNTER COUNT
seconds.until.enterprise.license.expiry
Seconds until enterprise license expiry (0 if no license present or running without enterprise features) GAUGE TIMESTAMP_SEC
security.certificate.expiration.ca
Expiration for the CA certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.ca-client-tenant
Expiration for the Tenant Client CA certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.client
Minimum expiration for client certificates, labeled by SQL user. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.client-ca
Expiration for the client CA certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.client-tenant
Expiration for the Tenant Client certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.node
Expiration for the node certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.node-client
Expiration for the node's client certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.ui
Expiration for the UI certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
security.certificate.expiration.ui-ca
Expiration for the UI CA certificate. 0 means no certificate or error. GAUGE TIMESTAMP_SEC
sql.conn.failures
Number of SQL connection failures COUNTER COUNT
sql.conn.latency
Latency to establish and authenticate a SQL connection HISTOGRAM NANOSECONDS
sql.conns
Number of open SQL connections GAUGE COUNT
sql.ddl.count
Number of SQL DDL statements successfully executed COUNTER COUNT
sql.delete.count
Number of SQL DELETE statements successfully executed COUNTER COUNT
sql.distsql.contended_queries.count
Number of SQL queries that experienced contention COUNTER COUNT
sql.exec.latency
Latency of SQL statement execution HISTOGRAM NANOSECONDS
sql.failure.count
Number of statements resulting in a planning or runtime error COUNTER COUNT
sql.full.scan.count
Number of full table or index scans COUNTER COUNT
sql.insert.count
Number of SQL INSERT statements successfully executed COUNTER COUNT
sql.mem.root.current
Current sql statement memory usage for root GAUGE BYTES
sql.new_conns
Number of SQL connections created COUNTER COUNT
sql.query.count
Number of SQL operations started including queries, and transaction control statements COUNTER COUNT
sql.routine.delete.count
Number of SQL DELETE statements successfully executed within routine invocation COUNTER COUNT
sql.routine.insert.count
Number of SQL INSERT statements successfully executed within routine invocation COUNTER COUNT
sql.routine.select.count
Number of SQL SELECT statements successfully executed within routine invocation COUNTER COUNT
sql.routine.update.count
Number of SQL UPDATE statements successfully executed within routine invocation COUNTER COUNT
sql.select.count
Number of SQL SELECT statements successfully executed COUNTER COUNT
sql.service.latency
Latency of SQL request execution HISTOGRAM NANOSECONDS
sql.statements.active
Number of currently active user SQL statements GAUGE COUNT
sql.txn.abort.count
Number of SQL transaction abort errors COUNTER COUNT
sql.txn.begin.count
Number of SQL transaction BEGIN statements successfully executed COUNTER COUNT
sql.txn.commit.count
Number of SQL transaction COMMIT statements successfully executed COUNTER COUNT
sql.txn.latency
Latency of SQL transactions HISTOGRAM NANOSECONDS
sql.txn.rollback.count
Number of SQL transaction ROLLBACK statements successfully executed COUNTER COUNT
sql.txns.open
Number of currently open user SQL transactions GAUGE COUNT
sql.update.count
Number of SQL UPDATE statements successfully executed COUNTER COUNT
storage.disk-slow
Number of instances of disk operations taking longer than 10s COUNTER COUNT
storage.disk-stalled
Number of instances of disk operations taking longer than 20s COUNTER COUNT
storage.disk.iopsinprogress
IO operations currently in progress on the store's disk (as reported by the OS) GAUGE COUNT
storage.disk.read-max.bytespersecond
Maximum rate at which bytes were read from disk (as reported by the OS) GAUGE BYTES
storage.disk.read.bytes
Bytes read from the store's disk since this process started (as reported by the OS) COUNTER BYTES
storage.disk.read.count
Disk read operations on the store's disk since this process started (as reported by the OS) COUNTER COUNT
storage.disk.write-max.bytespersecond
Maximum rate at which bytes were written to disk (as reported by the OS) GAUGE BYTES
storage.disk.write.bytes
Bytes written to the store's disk since this process started (as reported by the OS) COUNTER BYTES
storage.disk.write.count
Disk write operations on the store's disk since this process started (as reported by the OS) COUNTER COUNT
storage.keys.tombstone.count
Approximate count of DEL, SINGLEDEL and RANGEDEL internal keys across the storage engine. GAUGE COUNT
storage.l0-level-size
Size of the SSTables in level 0 GAUGE BYTES
storage.wal.failover.switch.count
Count of the number of times WAL writing has switched from primary to secondary and vice versa. COUNTER COUNT
storage.wal.failover.write_and_sync.latency
The observed latency for writing and syncing to the logical Write-Ahead Log. HISTOGRAM NANOSECONDS
storage.wal.fsync.latency
The fsync latency to the Write-Ahead Log device. HISTOGRAM NANOSECONDS
storage.write-stall-nanos
Total write stall duration in nanos COUNTER NANOSECONDS
storage.write-stalls
Number of instances of intentional write stalls to backpressure incoming writes COUNTER COUNT
storeliveness.heartbeat.failures
Number of Store Liveness heartbeats that failed to be sent out by the Store Liveness Support Manager COUNTER COUNT
sys.cgo.allocbytes
Current bytes of memory allocated by cgo GAUGE BYTES
sys.cgo.totalbytes
Total bytes of memory allocated by cgo, but not released GAUGE BYTES
sys.cpu.combined.percent-normalized
Current user+system cpu percentage consumed by the CRDB process, normalized 0-1 by number of cores GAUGE PERCENT
sys.cpu.host.combined.percent-normalized
Current user+system cpu percentage across the whole machine, normalized 0-1 by number of cores GAUGE PERCENT
sys.cpu.sys.ns
Total system cpu time consumed by the CRDB process COUNTER NANOSECONDS
sys.cpu.sys.percent
Current system cpu percentage consumed by the CRDB process GAUGE PERCENT
sys.cpu.user.ns
Total user cpu time consumed by the CRDB process COUNTER NANOSECONDS
sys.cpu.user.percent
Current user cpu percentage consumed by the CRDB process GAUGE PERCENT
sys.gc.count
Total number of GC runs COUNTER COUNT
sys.gc.pause.ns
Total GC pause COUNTER NANOSECONDS
sys.gc.pause.percent
Current GC pause percentage GAUGE PERCENT
sys.go.allocbytes
Current bytes of memory allocated by go GAUGE BYTES
sys.go.heap.allocbytes
Cumulative bytes allocated for heap objects. COUNTER BYTES
sys.go.heap.heapfragmentbytes
Total heap fragmentation bytes, derived from bytes in in-use spans minus bytes allocated GAUGE BYTES
sys.go.totalbytes
Total bytes of memory allocated by go, but not released GAUGE BYTES
sys.goroutines
Current number of goroutines GAUGE COUNT
sys.host.disk.iopsinprogress
IO operations currently in progress on this host (as reported by the OS) GAUGE COUNT
sys.host.disk.read.bytes
Bytes read from all disks since this process started (as reported by the OS) COUNTER BYTES
sys.host.disk.read.count
Disk read operations across all disks since this process started (as reported by the OS) COUNTER COUNT
sys.host.disk.write.bytes
Bytes written to all disks since this process started (as reported by the OS) COUNTER BYTES
sys.host.disk.write.count
Disk write operations across all disks since this process started (as reported by the OS) COUNTER COUNT
sys.host.net.recv.bytes
Bytes received on all network interfaces since this process started (as reported by the OS) COUNTER BYTES
sys.host.net.send.bytes
Bytes sent on all network interfaces since this process started (as reported by the OS) COUNTER BYTES
sys.host.net.send.tcp.retrans_segs
The number of TCP segments retransmitted across all network interfaces. This can indicate packet loss occurring in the network. However, it can also be caused by recipient nodes not consuming packets in a timely manner, or the local node overflowing its outgoing buffers, for example due to overload.

Retransmissions also occur in the absence of problems, as modern TCP stacks err on the side of aggressively retransmitting segments.

The linux tool 'ss -i' can show the Linux kernel's smoothed view of round-trip latency and variance on a per-connection basis. Additionally, 'netstat -s' shows all TCP counters maintained by the kernel.

COUNTER COUNT
sys.rss
Current process RSS GAUGE BYTES
sys.runnable.goroutines.per.cpu
Average number of goroutines that are waiting to run, normalized by number of cores GAUGE COUNT
sys.totalmem
Total memory (both free and used) GAUGE BYTES
sys.uptime
Process uptime COUNTER SECONDS
txn.restarts
Number of restarted KV transactions HISTOGRAM COUNT
txn.restarts.serializable
Number of restarts due to a forwarded commit timestamp and isolation=SERIALIZABLE COUNTER COUNT
txn.restarts.txnaborted
Number of restarts due to an abort by a concurrent transaction (usually due to deadlock) COUNTER COUNT
txn.restarts.txnpush
Number of restarts due to a transaction push failure COUNTER COUNT
txn.restarts.unknown
Number of restarts due to a unknown reasons COUNTER COUNT
txn.restarts.writetooold
Number of restarts due to a concurrent writer committing first COUNTER COUNT
txnwaitqueue.deadlocks_total
Number of deadlocks detected by the txn wait queue COUNTER COUNT

Total metrics (216)

See also

×