Egor Kuts created IGNITE-27500:
----------------------------------
Summary: Extend metrics message in the logs
Key: IGNITE-27500
URL: https://issues.apache.org/jira/browse/IGNITE-27500
Project: Ignite
Issue Type: Task
Reporter: Egor Kuts
Assignee: Egor Kuts
in AI3 there is a metrics message printed every minute
2025-11-10 08:20:40:198 +0000
[INFO][%node1%metrics-exporter-logPush-0][LogPushExporter] Metric report:
metastorage [IdempotentCacheSize=5, SafeTimeLag=156]
placement-driver [ActiveLeasesCount=0, CurrentPendingAssignmentsSize=0,
CurrentStableAssignmentsSize=25, LeasesCreated=200, LeasesProlonged=136257,
LeasesPublished=75, LeasesWithoutCandidates=0]
resource.vacuum [MarkedForVacuumTransactionMetaCount=1,
SkippedForFurtherProcessingUnfinishedTransactionCount=0,
VacuumizedPersistentTransactionMetaCount=0, VacuumizedVolatileTxnMetaCount=1]
We need to improve this message:
# Change the log format to oneliner.
# Replace prefix 'Metric report:' to 'Metrics for local node'
# Add uptime of a node(jvm process), state=init/not(if possible to determine)
# Add off-heap memory
# Add node-id(consistent and ephimerial)
# Cluster physical topology
# Brief network info(same as in ignite-2)
# CPU usage
# Support mission crytical thread pools' metrics(for now artitionoperation
only), but make it extensible
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
