Hi,
I'm looking to improve performance and avoid the subsearch_maxout issue with a join on two source types.
I'm joining on a one to many field 'Correlator', srctypeA has will have one unique entry for 'Correlator' and srctypeB has multiple entries. So when my search on srctypeA returns a value for Correlator I want to timechart by a field in records with matching Correlator in srctypeB.
So the current search below works, I'm looking for better performing options if anyone has any suggestions.
index=index1 sourcetype=srctypeA field1=ABC*
| join max=0 type=inner Correlator [ search index=index1 sourcetype=srctypeB ]
| timechart span=1d count by field2
Hi
I hope that this will help you: https://community.splunk.com/t5/Splunk-Search/What-is-the-relation-between-the-Splunk-inner-left-joi...
r. Ismo
thanks,
thats is what I have been attempting and I get to the stats point ok with 121k events returned but trying to get the data into a timechart is eluding me at the moment.
Join has those limits (size of result set and timeout), which probably hits in your case.
Have you try Woodcock’s inner join example with stats and where? I suppose that it should help you.