Two indexes are failing bundle validation checks on my cluster master with this error message:
[Critical] App='system' with replicated index='_introspection' is neither in the bundle downloaded from master nor managed by local deployment client. Either define this index at the master or specify repFactor=0 on peer to skip replication.
[Critical] App='system' with replicated index='_metrics' is neither in the bundle downloaded from master nor managed by local deployment client. Either define this index at the master or specify repFactor=0 on peer to skip replication.
The message is sent to the cluster master from the search peers. The message means that there are indexes on the peer nodes that are set for replication, but are not defined in the cluster master's configuration bundle.
The indexes are flagged for replication because the repFactor = auto
. This setting is not set in system/default
out of the box so it's likely set in other apps, perhaps as part of a default
stanza.
You can find out for sure by running the btool command $SPLUNK_HOME/bin/splunk btool indexes list _metrics --debug | grep repFactor
where _metrics
can also be _introspection
. The result will show only the configuration file where repFactor
is defined.
Two potential work-arounds are:
To get the validation to pass, add a definition of the indexes that are throwing the error to the cluster master configuration bundle. For example, add these stanzas to $SPLUNK_HOME/master-apps/_cluster/local/indexes.conf
on the master node and push the bundle again:
[_metrics]
repFactor = 0
[_introspection]
repFactor = 0
The cluster bundle can be forced to skip validated on the command line interface. The entire validation functionality protects against breaking indexers or losing data with faulty configuration, therefore it is discouraged to circumvent this protection given the ease of the other workaround options available for this. Learn more about validation and how to skip it, as a last resort, at Update common peer configurations and apps in the Managing Indexers and Clusters of Indexers documentation.
Learn more about the repFactor setting in Configure the peer indexes in an indexer cluster of the Managing Indexers and Clusters of Indexers documentation.
Learn more about using btool at Use btool to troubleshoot configurations within the Troubleshooting Manual.
The message is sent to the cluster master from the search peers. The message means that there are indexes on the peer nodes that are set for replication, but are not defined in the cluster master's configuration bundle.
The indexes are flagged for replication because the repFactor = auto
. This setting is not set in system/default
out of the box so it's likely set in other apps, perhaps as part of a default
stanza.
You can find out for sure by running the btool command $SPLUNK_HOME/bin/splunk btool indexes list _metrics --debug | grep repFactor
where _metrics
can also be _introspection
. The result will show only the configuration file where repFactor
is defined.
Two potential work-arounds are:
To get the validation to pass, add a definition of the indexes that are throwing the error to the cluster master configuration bundle. For example, add these stanzas to $SPLUNK_HOME/master-apps/_cluster/local/indexes.conf
on the master node and push the bundle again:
[_metrics]
repFactor = 0
[_introspection]
repFactor = 0
The cluster bundle can be forced to skip validated on the command line interface. The entire validation functionality protects against breaking indexers or losing data with faulty configuration, therefore it is discouraged to circumvent this protection given the ease of the other workaround options available for this. Learn more about validation and how to skip it, as a last resort, at Update common peer configurations and apps in the Managing Indexers and Clusters of Indexers documentation.
Learn more about the repFactor setting in Configure the peer indexes in an indexer cluster of the Managing Indexers and Clusters of Indexers documentation.
Learn more about using btool at Use btool to troubleshoot configurations within the Troubleshooting Manual.