Is there any way to re-create the TSIDX files by journal.gz?
Previously we had enabled the walls:
enabletsidxredctuon = 1
timePeriodInSecBeforeTsidxReduction = 2592000
And now we're having problems with searches that need to fetch data older than 1 month. returning the following alert:
"Search on most recent data has completed. Expect slower search sppeds as we search teh reduced buckets."
We understand that by the configuration we did we ended up deleting the older TSIDX, but now we need a way to recreate them. It's possible?
After you restore the settings to not do it anymore the steps to perform to rebuild the full tsidx buckets is documented: https://docs.splunk.com/Documentation/Splunk/7.1.1/Indexer/Reducetsidxdiskusage#Restore_reduced_buck...
I read the KB:
http://docs.splunk.com/Documentation/Splunk/7.1.1/Indexer/Bucketissues#Rebuild_a_single_bucket
and with the knowledge acquired I followed the following steps to solve the problem:
I stopped Splunk
I invoked the following command:
splunk fsck repair --all-buckets-one-index --index-name = main
After 4 hours the command ended I started Splunk again and the problem had been solved.
Thanks a lot for the help!
After you restore the settings to not do it anymore the steps to perform to rebuild the full tsidx buckets is documented: https://docs.splunk.com/Documentation/Splunk/7.1.1/Indexer/Reducetsidxdiskusage#Restore_reduced_buck...
Unfortunately I ran the command in all bucket that was missing the .rbsentinel file, done that I started Splunk again but the problem of slowness for older dates still persists. Is there anything else that can be done?
thanks for the help @acharlieh
I've been on the path: / opt / splunk / var / lib / splunk / defaultdb / db
And I ran the command on every bucket that did not have a ..rbsentinel file
After the executed command the files have been recreated but the slowness continues.