Getting Data In

Per index archiving

Path Finder

If I define this in .../local/indexes.conf


[default]
coldToFrozenDir = $SPLUNK_DB/frozenArchive

Will Splunk roll "frozen" buckets from all indexes inside var/lib/splunk/frozenArchive? If so, can this cause conflicts or is there auto-suffix/prefix for file names? My idea was to have a general coldToFrozenDir rather than per index. This would (in theory) take care of new indexes.

1 Solution

Splunk Employee
Splunk Employee

I'm not sure this will work as you want, but I believe that if you simply put coldToFrozenDir = $SPLUNK_DB/frozenArchive in the indexes.conf file at the top, not under a stanza header, it will have the effect you're looking for. However, please note that once the buckets are rolled into the directory, there is nothing that will tell you what index it originally came from. They will still work and still be usable, but the index name will be lost.

View solution in original post

Path Finder

it would still be nice to have a $SPUNK_INDEX variable so that one could add a generic coldToFrozenDir var

0 Karma

Splunk Employee
Splunk Employee

I'm not sure this will work as you want, but I believe that if you simply put coldToFrozenDir = $SPLUNK_DB/frozenArchive in the indexes.conf file at the top, not under a stanza header, it will have the effect you're looking for. However, please note that once the buckets are rolled into the directory, there is nothing that will tell you what index it originally came from. They will still work and still be usable, but the index name will be lost.

View solution in original post

Path Finder

Perhaps there is a special variable for index names?
e.g.
coldToFrozenDir = $SPLUNKDB/frozenArchive/$SPLUNKINDEX
then the structure would be:
.../frozenArchive/main
.../frozenArchive/_audit
...[etc]

0 Karma