open-consul/website/source/docs/agent/telemetry.html.md

28 KiB

layout page_title sidebar_current description
docs Telemetry docs-agent-telemetry The Consul agent collects various runtime metrics about the performance of different libraries and subsystems. These metrics are aggregated on a ten second interval and are retained for one minute.

Telemetry

The Consul agent collects various runtime metrics about the performance of different libraries and subsystems. These metrics are aggregated on a ten second interval and are retained for one minute.

To view this data, you must send a signal to the Consul process: on Unix, this is USR1 while on Windows it is BREAK. Once Consul receives the signal, it will dump the current telemetry information to the agent's stderr.

This telemetry information can be used for debugging or otherwise getting a better view of what Consul is doing.

Additionally, if the telemetry configuration options are provided, the telemetry information will be streamed to a statsite or statsd server where it can be aggregated and flushed to Graphite or any other metrics store. This information can also be viewed with the metrics endpoint in JSON format or using Prometheus format.

Below is sample output of a telemetry dump:

[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.num_goroutines': 19.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.alloc_bytes': 755960.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.malloc_count': 7550.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.free_count': 4387.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.heap_objects': 3163.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.total_gc_pause_ns': 1151002.000
[2014-01-29 10:56:50 -0800 PST][G] 'consul-agent.runtime.total_gc_runs': 4.000
[2014-01-29 10:56:50 -0800 PST][C] 'consul-agent.agent.ipc.accept': Count: 5 Sum: 5.000
[2014-01-29 10:56:50 -0800 PST][C] 'consul-agent.agent.ipc.command': Count: 10 Sum: 10.000
[2014-01-29 10:56:50 -0800 PST][C] 'consul-agent.serf.events': Count: 5 Sum: 5.000
[2014-01-29 10:56:50 -0800 PST][C] 'consul-agent.serf.events.foo': Count: 4 Sum: 4.000
[2014-01-29 10:56:50 -0800 PST][C] 'consul-agent.serf.events.baz': Count: 1 Sum: 1.000
[2014-01-29 10:56:50 -0800 PST][S] 'consul-agent.memberlist.gossip': Count: 50 Min: 0.007 Mean: 0.020 Max: 0.041 Stddev: 0.007 Sum: 0.989
[2014-01-29 10:56:50 -0800 PST][S] 'consul-agent.serf.queue.Intent': Count: 10 Sum: 0.000
[2014-01-29 10:56:50 -0800 PST][S] 'consul-agent.serf.queue.Event': Count: 10 Min: 0.000 Mean: 2.500 Max: 5.000 Stddev: 2.121 Sum: 25.000

Key Metrics

When telemetry is being streamed to an external metrics store, the interval is defined to be that store's flush interval. Otherwise, the interval can be assumed to be 10 seconds when retrieving metrics from the built-in store using the above described signals.

Agent Health

These metrics are used to monitor the health of specific Consul agents.

Metric Description Unit Type
`consul.client.rpc` This increments whenever a Consul agent in client mode makes an RPC request to a Consul server. This gives a measure of how much a given agent is loading the Consul servers. Currently, this is only generated by agents in client mode, not Consul servers. requests counter
`consul.client.rpc.exceeded` This increments whenever a Consul agent in client mode makes an RPC request to a Consul server gets rate limited by that agent's [`limits`](/docs/agent/options.html#limits) configuration. This gives an indication that there's an abusive application making too many requests on the agent, or that the rate limit needs to be increased. Currently, this only applies to agents in client mode, not Consul servers. rejected requests counter
`consul.client.rpc.failed` This increments whenever a Consul agent in client mode makes an RPC request to a Consul server and fails. requests counter
`consul.client.api.catalog_register.` This increments whenever a Consul agent receives a catalog register request. requests counter
`consul.client.api.success.catalog_register.` This increments whenever a Consul agent successfully responds to a catalog register request. requests counter
`consul.client.rpc.error.catalog_register.` This increments whenever a Consul agent receives an RPC error for a catalog register request. errors counter
`consul.client.api.catalog_deregister.` This increments whenever a Consul agent receives a catalog de-register request. requests counter
`consul.client.api.success.catalog_deregister.` This increments whenever a Consul agent successfully responds to a catalog de-register request. requests counter
`consul.client.rpc.error.catalog_deregister.` This increments whenever a Consul agent receives an RPC error for a catalog de-register request. errors counter
`consul.client.api.catalog_datacenters.` This increments whenever a Consul agent receives a request to list datacenters in the catalog. requests counter
`consul.client.api.success.catalog_datacenters.` This increments whenever a Consul agent successfully responds to a request to list datacenters. requests counter
`consul.client.rpc.error.catalog_datacenters.` This increments whenever a Consul agent receives an RPC error for a request to list datacenters. errors counter
`consul.client.api.catalog_nodes.` This increments whenever a Consul agent receives a request to list nodes from the catalog. requests counter
`consul.client.api.success.catalog_nodes.` This increments whenever a Consul agent successfully responds to a request to list nodes. requests counter
`consul.client.rpc.error.catalog_nodes.` This increments whenever a Consul agent receives an RPC error for a request to list nodes. errors counter
`consul.client.api.catalog_services.` This increments whenever a Consul agent receives a request to list services from the catalog. requests counter
`consul.client.api.success.catalog_services.` This increments whenever a Consul agent successfully responds to a request to list services. requests counter
`consul.client.rpc.error.catalog_services.` This increments whenever a Consul agent receives an RPC error for a request to list services. errors counter
`consul.client.api.catalog_service_nodes.` This increments whenever a Consul agent receives a request to list nodes offering a service. requests counter
`consul.client.api.success.catalog_service_nodes.` This increments whenever a Consul agent successfully responds to a request to list nodes offering a service. requests counter
`consul.client.rpc.error.catalog_service_nodes.` This increments whenever a Consul agent receives an RPC error for a request to list nodes offering a service. errors counter
`consul.client.api.catalog_node_services.` This increments whenever a Consul agent receives a request to list services registered in a node. requests counter
`consul.client.api.success.catalog_node_services.` This increments whenever a Consul agent successfully responds to a request to list services in a service. requests counter
`consul.client.rpc.error.catalog_node_services.` This increments whenever a Consul agent receives an RPC error for a request to list services in a service. errors counter
`consul.runtime.num_goroutines` This tracks the number of running goroutines and is a general load pressure indicator. This may burst from time to time but should return to a steady state value. number of goroutines gauge
`consul.runtime.alloc_bytes` This measures the number of bytes allocated by the Consul process. This may burst from time to time but should return to a steady state value. bytes gauge
`consul.runtime.heap_objects` This measures the number of objects allocated on the heap and is a general memory pressure indicator. This may burst from time to time but should return to a steady state value. number of objects gauge
`consul.acl.cache_hit` The number of ACL cache hits. hits counter
`consul.acl.cache_miss` The number of ACL cache misses. misses counter
`consul.acl.replication_hit` The number of ACL replication cache hits (when not running in the ACL datacenter). hits counter
`consul.dns.stale_queries` This increments when an agent serves a query within the allowed stale threshold. queries counter
`consul.dns.ptr_query.` This measures the time spent handling a reverse DNS query for the given node. ms timer
`consul.dns.domain_query.` This measures the time spent handling a domain query for the given node. ms timer
`consul.http..` This tracks how long it takes to service the given HTTP request for the given verb and path. Paths do not include details like service or key names, for these an underscore will be present as a placeholder (eg. `consul.http.GET.v1.kv._`) ms timer

Server Health

These metrics are used to monitor the health of the Consul servers.

Metric Description Unit Type
`consul.raft.state.leader` This increments whenever a Consul server becomes a leader. If there are frequent leadership changes this may be indication that the servers are overloaded and aren't meeting the soft real-time requirements for Raft, or that there are networking problems between the servers. leadership transitions / interval counter
`consul.raft.state.candidate` This increments whenever a Consul server starts an election. If this increments without a leadership change occurring it could indicate that a single server is overloaded or is experiencing network connectivity issues. election attempts / interval counter
`consul.raft.apply` This counts the number of Raft transactions occurring over the interval, which is a general indicator of the write load on the Consul servers. raft transactions / interval counter
`consul.raft.commitTime` This measures the time it takes to commit a new entry to the Raft log on the leader. ms timer
`consul.raft.leader.dispatchLog` This measures the time it takes for the leader to write log entries to disk. ms timer
`consul.raft.replication.appendEntries` This measures the time it takes to replicate log entries to followers. This is a general indicator of the load pressure on the Consul servers, as well as the performance of the communication between the servers. ms timer
`consul.raft.leader.lastContact` This will only be emitted by the Raft leader and measures the time since the leader was last able to contact the follower nodes when checking its leader lease. It can be used as a measure for how stable the Raft timing is and how close the leader is to timing out its lease.

The lease timeout is 500 ms times the [`raft_multiplier` configuration](/docs/agent/options.html#raft_multiplier), so this telemetry value should not be getting close to that configured value, otherwise the Raft timing is marginal and might need to be tuned, or more powerful servers might be needed. See the [Server Performance](/docs/guides/performance.html) guide for more details.
ms timer
`consul.acl.apply` This measures the time it takes to complete an update to the ACL store. ms timer
`consul.acl.fault` This measures the time it takes to fault in the rules for an ACL during a cache miss. ms timer
`consul.acl.fetchRemoteACLs` This measures the time it takes to fetch remote ACLs during replication. ms timer
`consul.acl.updateLocalACLs` This measures the time it takes to apply replication changes to the local ACL store. ms timer
`consul.acl.replicateACLs` This measures the time it takes to do one pass of the ACL replication algorithm. ms timer
`consul.acl.resolveToken` This measures the time it takes to resolve an ACL token. ms timer
`consul.rpc.accept_conn` This increments when a server accepts an RPC connection. connections counter
`consul.catalog.register` This measures the time it takes to complete a catalog register operation. ms timer
`consul.catalog.deregister` This measures the time it takes to complete a catalog deregister operation. ms timer
`consul.fsm.register` This measures the time it takes to apply a catalog register operation to the FSM. ms timer
`consul.fsm.deregister` This measures the time it takes to apply a catalog deregister operation to the FSM. ms timer
`consul.fsm.acl.` This measures the time it takes to apply the given ACL operation to the FSM. ms timer
`consul.fsm.session.` This measures the time it takes to apply the given session operation to the FSM. ms timer
`consul.fsm.kvs.` This measures the time it takes to apply the given KV operation to the FSM. ms timer
`consul.fsm.tombstone.` This measures the time it takes to apply the given tombstone operation to the FSM. ms timer
`consul.fsm.coordinate.batch-update` This measures the time it takes to apply the given batch coordinate update to the FSM. ms timer
`consul.fsm.prepared-query.` This measures the time it takes to apply the given prepared query update operation to the FSM. ms timer
`consul.fsm.txn` This measures the time it takes to apply the given transaction update to the FSM. ms timer
`consul.fsm.autopilot` This measures the time it takes to apply the given autopilot update to the FSM. ms timer
`consul.fsm.persist` This measures the time it takes to persist the FSM to a raft snapshot. ms timer
`consul.kvs.apply` This measures the time it takes to complete an update to the KV store. ms timer
`consul.leader.barrier` This measures the time spent waiting for the raft barrier upon gaining leadership. ms timer
`consul.leader.reconcile` This measures the time spent updating the raft store from the serf member information. ms timer
`consul.leader.reconcileMember` This measures the time spent updating the raft store for a single serf member's information. ms timer
`consul.leader.reapTombstones` This measures the time spent clearing tombstones. ms timer
`consul.prepared-query.apply` This measures the time it takes to apply a prepared query update. ms timer
`consul.prepared-query.explain` This measures the time it takes to process a prepared query explain request. ms timer
`consul.prepared-query.execute` This measures the time it takes to process a prepared query execute request. ms timer
`consul.prepared-query.execute_remote` This measures the time it takes to process a prepared query execute request that was forwarded to another datacenter. ms timer
`consul.rpc.raft_handoff` This increments when a server accepts a Raft-related RPC connection. connections counter
`consul.rpc.request_error` This increments when a server returns an error from an RPC request. errors counter
`consul.rpc.request` This increments when a server receives a Consul-related RPC request. requests counter
`consul.rpc.query` This increments when a server receives a (potentially blocking) RPC query. queries counter
`consul.rpc.cross-dc` This increments when a server receives a (potentially blocking) cross datacenter RPC query. queries counter
`consul.rpc.consistentRead` This measures the time spent confirming that a consistent read can be performed. ms timer
`consul.session.apply` This measures the time spent applying a session update. ms timer
`consul.session.renew` This measures the time spent renewing a session. ms timer
`consul.session_ttl.invalidate` This measures the time spent invalidating an expired session. ms timer
`consul.txn.apply` This measures the time spent applying a transaction operation. ms timer
`consul.txn.read` This measures the time spent returning a read transaction. ms timer

Cluster Health

These metrics give insight into the health of the cluster as a whole.

Metric Description Unit Type
`consul.memberlist.msg.suspect` This increments when an agent suspects another as failed when executing random probes as part of the gossip protocol. These can be an indicator of overloaded agents, network problems, or configuration errors where agents can not connect to each other on the [required ports](/docs/agent/options.html#ports). suspect messages received / interval counter
`consul.serf.member.flap` Available in Consul 0.7 and later, this increments when an agent is marked dead and then recovers within a short time period. This can be an indicator of overloaded agents, network problems, or configuration errors where agents can not connect to each other on the [required ports](/docs/agent/options.html#ports). flaps / interval counter
`consul.serf.events` This increments when an agent processes an [event](/docs/commands/event.html). Consul uses events internally so there may be additional events showing in telemetry. There are also a per-event counters emitted as `consul.serf.events.`. events / interval counter
`consul.autopilot.failure_tolerance` This tracks the number of voting servers that the cluster can lose while continuing to function. servers gauge
`consul.autopilot.healthy` This tracks the overall health of the local server cluster. If all servers are considered healthy by Autopilot, this will be set to 1. If any are unhealthy, this will be 0. boolean gauge
`consul.session_ttl.active` This tracks the active number of sessions being tracked. sessions gauge
`consul.catalog.service.query.` This increments for each catalog query for the given service. queries counter
`consul.catalog.service.query-tag..` This increments for each catalog query for the given service with the given tag. queries counter
`consul.catalog.service.not-found.` This increments for each catalog query where the given service could not be found. queries counter
`consul.health.service.query.` This increments for each health query for the given service. queries counter
`consul.health.service.query-tag..` This increments for each health query for the given service with the given tag. queries counter
`consul.health.service.not-found.` This increments for each health query where the given service could not be found. queries counter

Connect Built-in Proxy Metrics

Consul Connect's built-in proxy is by default configured to log metrics to the same sink as the agent that starts it when running as a managed proxy.

When running in this mode it emits some basic metrics. These will be expanded upon in the future.

All metrics are prefixed with consul.proxy.<proxied-service-id> to distinguish between multiple proxies on a given host. The table below use web as an example service name for brevity.

Labels

Most labels have a dst label and some have a src label. When using metrics sinks and timeseries stores that support labels or tags, these allow aggregating the connections by service name.

Assuming all services are using a managed built-in proxy, you can get a complete overview of both number of open connections and bytes sent and recieved between all services by aggregating over these metrics.

For example aggregating over all upstream (i.e. outbound) connections which have both src and dst labels, you can get a sum of all the bandwidth in and out of a given service or the total number of connections between two services.

Metrics Reference

The standard go runtime metrics are exported by go-metrics as with Consul agent. The table below describes the additional metrics exported by the proxy.

Metric Description Unit Type
`consul.proxy.web.runtime.*` The same go runtime metrics as documented for the agent above. mixed mixed
`consul.proxy.web.inbound.conns` Shows the current number of connections open from inbound requests to the proxy. Where supported a `dst` label is added indicating the service name the proxy represents. connections gauge
`consul.proxy.web.inbound.rx_bytes` This increments by the number of bytes received from an inbound client connection. Where supported a `dst` label is added indicating the service name the proxy represents. bytes counter
`consul.proxy.web.inbound.tx_bytes` This increments by the number of bytes transfered to an inbound client connection. Where supported a `dst` label is added indicating the service name the proxy represents. bytes counter
`consul.proxy.web.upstream.conns` Shows the current number of connections open from a proxy instance to an upstream. Where supported a `src` label is added indicating the service name the proxy represents, and a `dst` label is added indicating the service name the upstream is connecting to. connections gauge
`consul.proxy.web.inbound.rx_bytes` This increments by the number of bytes received from an upstream connection. Where supported a `src` label is added indicating the service name the proxy represents, and a `dst` label is added indicating the service name the upstream is connecting to. bytes counter
`consul.proxy.web.inbound.tx_bytes` This increments by the number of bytes transfered to an upstream connection. Where supported a `src` label is added indicating the service name the proxy represents, and a `dst` label is added indicating the service name the upstream is connecting to. bytes counter