So I have some data for some reason that did not get index in my monitored filepath. I have a feeling it has something to do with the service writing to the file. It stopped writing for some time and then we restarted it and I updated the monitor to go to a new sourcetype in the same index.
It seems the data from 1/30 to today isn't indexed, maybe since the crc didn't change?
Is there a way to reindex the data from 2/1 to today without reindexing the other files?(Ex 1/30 e4ct.)
CAn I change the crc and add ignoreOlderThan?
Thanks for the help
For this single use case, you can use
crcSalt = <SOURCE> in inputs.conf to reindex only those files you believe need to be reindexed. You could add an entry in inputs.conf like:
[monitor://tmpdirectory] crcSalt = <SOURCE> index = whatever sourcetype = whatever
tmpdirectory to a location that you will copy the previously unindexed files into.
crcSalt = <SOURCE> in inputs.conf will tell splunk that the filename is what makes these files unique, and therefore it will index them even though it may think it has seen them before.
The above is a one time workaround, it is not intended to be used for your normal log rotation monitoring.
Good answer, however the root cause was me not configuring timestamp recognition correctly.
It seems that the field we were using for timestamp doesn't change, even when a event is updated we were still looking at one date instead of a different one, so the data from those files were there just under a different date.