I am getting below error for '_introspection' index-
The percentage of small buckets (75%) created over the last hour is high and exceeded the red thresholds (50%) for index=_introspection, and possibly more indexes, on this indexer. At the time this alert fired, total buckets created=4, small buckets=3
I want to know why this is happening with _introspection index?
I can understand if I increased hot bucket count then error may get resolved but I would like to know why it is happening?
If you are on Linux, the default setting in Splunk is to forward/index all files at /opt/splunk/var/log/splunk/*
This becomes an issue if you have logrotate configured, and have not updated dir monitoring.
Meaning, logrotate will either rotate log files and append a timestamp, or compress the files to .gz, or both.
Out of the box, Splunk will index all versions of the same file because it see's them as "new".
Ensure that you are not indexing locally, whitelist .log files, and blacklist everything else.
---- An upvote would be appreciated and Accept Solution if it helps!
Have you check that you have correct time on all your nodes from where you are collecting events by UF, HF? Quite often this could mean that events from nodes comes with timestamps which are not fitting to Splunk's default time range for events in one bucket. If/when events timestamp is out of range then splunk close current host bucket after it has inserted event and create new one. r. Ismo