(This is a continuation of https://community.splunk.com/t5/Splunk-Search/Creating-a-search-that-looks-up-values-from-one-logfil...)
So what I'm trying to do is compare values across two different logs that have a unique ID (see above link for more info). The problem I have is that the one log runs on a cronjob every 5 minutes, while the other log only runs when required. I'm using the below code:
index=foo (sourcetype=test1 OR sourcetype=test2) host=A* (source="/data/stuff/logfile1.log" OR source="/data/stuff/logfile2.log")
| eval coalesce(lastupdate_direc,file1ID)
| stats values(lastupdate_time) as lastupdate_time, values(file1ID) as file1ID by host, ID
| eval int_time=strptime(lastupdate_time, "%F %H:%M")
| eval timenow=now()
| eval diff_new=timenow-int_time
| eval days_since=((diff_new-14400)/60/60/24)
| table lastupdate_time host name ID days_since
As I'm trying to be nice to my indexer, I'm only looking at the past 15 minutes (because of time drift between the servers) so I get multiple lastupdate_time entries in the table and because of that, the days_since field will be left blank. I've tried using chart latest of the values, but I get no results. Am I just not outputting to the correct visualization function or would something other than table be better?
Thanks!