Hi, I have a couple searches where the main search can be limited a fair amount, lets say the last 2 weeks, but I have a subsearch that requires searching across all time, albeit on a small dataset.
Currently I handle this as follows:
event=LOGIN | where _time < relative_time(now(),-14d) | join type=left userId [search source=SmallSource.csv | table userId userProperty] | where userProperty = X | ...
With the time picker set to All Time. The SmallSource contains some user Properties tied to creation date of the user, so I need it to search across All Time, but this slows down the main search which must now search every event across time despite only needing the last 2 weeks!
Is there a way I can make this more efficient? I'm fairly new to Splunk so I'm not sure if I can put some of this info in another search, like a report I can reference, or if there is another way to restrict time more efficiently than this.
you can specify the earliest/latest time modifiers in your subsearch and use the time picker for the main search.
event=LOGIN | join type=left userId [search source=SmallSource.csv earliest=0 latest=now | table userId userProperty] | where userProperty = X | ...
I did not realize the time picker on the main search would not restrict the sub-search as well regardless. Thanks for the tip!
glad to help. but just to note, anytime you specify earlier or latest in your search, it will override the whatever is the timepicker. So even if the time picker is set to Last 7 Days, this search would only run for the past 15 minutes
index=blah sourcetype=meh earliest=-15m