Skip to main content
Version: v1.5.x

Monitoring DORIS Database FE Monitoring

Collect and monitor general performance metrics for DORIS database FE. Supports DORIS 2.0.0.

Protocol: HTTP

Pre-monitoring operations

Check the fe/conf/fe.conf file to obtain the value of the http_port configuration item, which is used for monitoring.

Configuration Parameters

Parameter NameParameter Description
Monitor HostThe monitored target's IPV4, IPV6, or domain name. Note: Without the protocol header (e.g., https://, http://)
Task NameA unique name identifying this monitoring task
PortThe port provided by the database to the outside, default is 8030 ,get the value of thehttp_port configuration item
Query TimeoutThe timeout for the connection to not respond, in milliseconds, default is 3000 milliseconds
Database NameOptional database instance name
Collection IntervalThe interval between periodic data collections for monitoring, in seconds, with a minimum interval of 30 seconds
Probe RequiredWhether to probe and check the availability of monitoring before adding new monitoring, continue with the addition or modification operation only if the probe is successful
DescriptionAdditional notes and descriptions for this monitoring task

Collection Indicators

Metric Set: doris_fe_connection_total

Metric NameMetric UnitMetric help description
valueNoneThe current number of MySQL port connections on FE

Metric Set: doris_fe_edit_log_clean

Should not fail; if it does, manual intervention is required.

Metric NameMetric UnitMetric help description
successNoneThe number of successful cleanups of historical metadata logs
failedNoneThe number of failed cleanups of historical metadata logs

Metric Set: doris_fe_edit_log

Metric NameMetric UnitMetric help description
writeNoneThe count of metadata log write operations
readNoneThe count of metadata log read operations
currentNoneThe current number of metadata logs
accumulated_bytesBytesThe cumulative value of metadata log writes
current_bytesBytesThe current value of metadata logs

Metric Set: doris_fe_image_clean

Should not fail; if it does, manual intervention is required.

Metric NameMetric UnitMetric help description
successNoneThe number of successful cleanups of historical metadata image files
failedNoneThe number of failed cleanups of historical metadata image files

Metric Set: doris_fe_image_write

Should not fail; if it does, manual intervention is required.

Metric NameMetric UnitMetric help description
successNoneThe number of successful generations of metadata image files
failedNoneThe number of failed generations of metadata image files

Metric Set: doris_fe_query_err

Metric NameMetric UnitMetric help description
valueNoneThe cumulative value of erroneous queries

Metric Set: doris_fe_max_journal_id

Metric NameMetric UnitMetric help description
valueNoneThe current maximum metadata log ID on the FE node. If it is a Master FE, it is the maximum ID currently written; if it is a non-Master FE, it represents the maximum metadata log ID currently being replayed. Used to observe if there is a large gap between the IDs of multiple FEs. A large gap indicates issues with metadata synchronization

Metric Set: doris_fe_max_tablet_compaction_score

Metric NameMetric UnitMetric help description
valueNoneThe largest compaction score value among all BE nodes. This value can observe the current cluster's maximum compaction score to judge if it is too high. If too high, there may be delays in queries or writes

Metric Set: doris_fe_qps

Metric NameMetric UnitMetric help description
valueNoneThe number of queries per second on the current FE (only counts query requests)

Metric Set: doris_fe_query_err_rate

Metric NameMetric UnitMetric help description
valueNoneThe number of erroneous queries per second

Metric Set: doris_fe_report_queue_size

Metric NameMetric UnitMetric help description
valueNoneThe length of the queue for various regular reporting tasks on the BE side at the FE end. This value reflects the degree of blocking of reporting tasks on the Master FE node. A larger value indicates insufficient processing capacity on the FE

Metric Set: doris_fe_rps

Metric NameMetric UnitMetric help description
valueNoneThe number of requests per second on the current FE (includes queries and other types of statements)

Metric Set: doris_fe_scheduled_tablet_num

Metric NameMetric UnitMetric help description
valueNoneThe number of tablets currently being scheduled by the Master FE node. This includes tablets that are being repaired and tablets that are being balanced. This value can reflect the number of tablets currently migrating in the cluster. If there is a value for a long time, it indicates that the cluster is unstable

Metric Set: doris_fe_txn_status

Can observe the number of import transactions in various states to determine if there is a backlog.

Metric NameMetric UnitMetric help description
unknownNoneUnknown state
prepareNoneIn preparation
committedNoneCommitted
visibleNoneVisible
abortedNoneAborted / Revoked