We currently have
Events indexed Earliest event Latest event 452,254,458 07/23/2000 11:06:54 04/07/2011 11:04:07
we want to purge all but the last 7 days
I created a file in /opt/app/splunk/etc/system/local/
cat indexes.conf [main] frozenTimePeriodInSecs = 604800
How long should it take before I see the reults change in the SUmmary page
I refresh about every 2 minutes
I added
maxTotalDataSizeMB = 25000
to the /opt/app/splunk/etc/system/local/indexes.conf file still have the same problem
A bucket will only be purged when the newest event in the bucket is older than the frozenTimePeriodInSecs. That means that you need to configure your buckets such that they will have a relatively small amount of data (or a relatively small percentage of data).
By default, the main index buckets are pretty large (how big depends on the filesystem), but you can configure that to be, say, your daily data volume, and then on average you'll lose one bucket per day, which should contain events between exactly 7 and approximately 8 days old (depending on the particular week's data volumes).
Does this make sense?