Event-data within retention time is missing

[ufotech]

Hi

For a given index with retention of 91 days configured, we find some hosts having events for the full 91 days.
Some other hosts only retain 78 days and some other host even less than that.

This is a indexer-cluster-environment having 4 peers and a distributed indexer.conf.

Index:
[ix_windows_security_logs]
coldPath = $SPLUNK_DB/ix_windows_security_logs/colddb
enableTsidxReduction = 0
rtRouterThreads =
suspendHotRollByDeleteQuery = 0
enableOnlineBucketRepair = 1
frozenTimePeriodInSecs = 7862400
archiver.enableDataArchive = 0
bucketRebuildMemoryHint = 0
minHotIdleSecsBeforeForceRoll = 0
syncMeta = 1
maxTotalDataSizeMB = 716800
compressRawdata = 1
selfStorageThreads =
homePath = $SPLUNK_DB/ix_windows_security_logs/db
rtRouterQueueSize =
thawedPath = $SPLUNK_DB/ix_windows_security_logs/thaweddb
tsidxWritingLevel =

The index is not full at all:
ix_windows_security_logs Curent Size: 1 MB Max Size: 700 GB

Nor is the device:
/dev/mapper/vgappl-lvsplunk 3.1T 51G 2.9T 2% /opt/splunk

What could be the issue here?

Thank you

[ufotech · Accepted solution]

The Index-Size is 49 Gb. I found some default 'homePath.maxDataSizeMB' to be set to 51200. This may explain why data gets removed (freezed up) prematurely.

Cheers, Urs

[gcusello]

Hi @ufotech.

good for you.

Ciao and happy splunking.

Giuseppe

P.S.: Karma Points are appreciated 😉

[gcusello]

Hi @ufotech,

check if the index is working near to the maxTotalDataSizeMB (700 GB).

You can check it using the Splunk Monitoring Console.

This would explain why some events are discarded before retention time.

Ciao.

Giuseppe