Hive ACID metric properties for compaction observability

As an administrator, you must configure certain properties related to the Hive ACID metrics service to enable the collection of Hive ACID metrics that are required to display compaction-related alerts and charts in Cloudera Manager.

Basic properties

You must configure the following metrics to enable Hive ACID metrics collection:
hive.metastore.metrics.enabled
Default value: True
Enables or disables Hive metrics subsystem for the Hive Metastore (HMS) role.
hive.server2.metrics.enabled
Default value: True
Enables or disables Hive metrics subsystem for the HiveServer (HS2) role.
hive.metastore.acidmetrics.ext.on
Default value: False
Set this to property to True to collect additional acid related metrics outside of the Hive ACID metrics service. Enable this property for both HMS and HS2 roles.
metastore.acidmetrics.thread.on
Default value: True
Specifies whether to run Hive ACID related metrics collection on this metastore instance. Enable this property in only one HMS instance and disable the property in all other instances.

Advanced properties

Set these additional properties to configure how Hive ACID metrics service collects metric data:

The following properties have to be configured in the HMS instance where metastore.acidmetrics.thread.on is enabled:
metastore.acidmetrics.check.interval
Default value: 300 seconds
Specifies the time (in seconds) between Hive ACID metric collection jobs.
metastore.acidmetrics.table.aborted.txns.threshold
Default value: 1500
Hive metrics subsystem collects the number of tables that have a large number of aborted transactions. This property specifies the minimum number of aborted transactions required to consider a table.
The following properties have to be configured in the HS2 (Hive on Tez) service:
hive.txn.acid.metrics.cache.size
Default value: 100
Specifies the size of the ACID metrics cache, which is the maximum number of partitioned and unpartitioned tables with the most deltas that is included in the list of active, obsolete, and small deltas. Valid values are between 0 to 500.
hive.txn.acid.metrics.cache.duration
Default value: 7200 seconds
Specifies the maximum lifetime (in seconds) for an entry in the ACID metrics cache.
hive.txn.acid.metrics.reporting.interval
Default value: 30 seconds
Specifies the reporting period (in seconds) for ACID metrics.
hive.txn.acid.metrics.delta.num.threshold
Default value: 100
Specifies the minimum number of active delta files that a table or partition must have to be included in the ACID metrics report.
hive.txn.acid.metrics.obsolete.delta.num.threshold
Default value: 100
Specifies the minimum number of obsolete delta files that a table or partition must have to be included in the ACID metrics report.
hive.txn.acid.metrics.delta.check.threshold
Default value: 300 seconds
Specifies the minimum age (in seconds) for delta files to be included in the ACID metrics report. Delta files less than this age are not included in the report.
hive.txn.acid.metrics.delta.pct.threshold
Default value: 0.01
Specifies the percentage size of the delta files relative to the base directory. Delta files whose size is smaller than this threshold value are considered as small deltas and are not included in the ACID metrics report. (1 = 100%, so the default 0.01 = 1%)