topic Re: Filter records using time modifiers in Getting Data In

Filter records using time modifiers

dsiob — Thu, 27 Jul 2017 07:23:51 GMT

Can someone tell me why this is not working:-

I need to filter records having 'Start_Time' within the mentioned range:

Working:-

index=imt_mobile source="*LTS_Validation_1" |eval early=relative_time(now(),"-3w@w")|eval late=relative_time(now(),"-2w@w")| where Start_Time<late|table   "Track" "CO" early  Start_Time late|

Here Start_Time was before 'late' so coming fine.

Not Working:-

index=imt_mobile source="*LTS_Validation_1" |eval early=relative_time(now(),"-3w@w")|eval late=relative_time(now(),"-2w@w")| where Start_Time<"-2w@w"|table   "Track" "CO" early  Start_Time late|

Here it is not working!!

Re: Filter records using time modifiers

gcusello — Thu, 27 Jul 2017 08:31:15 GMT

Hi dsiob,
sorry but I don't understand your problem:
in a where condition you have to compare two similar fields: if you use " -2w@w " Splunk don't understand that you're speaking about a time in epochtime format, so you need to transform "-2w@w" in epochtime format to compare with StartTime.
To do this you have to use (as in your first example) |eval late=relative_time(now(),"-2w@w")| where Start_Time<late
So, what is the problem to use eval?
If instead you need to show in Human readable format the three dates use strftime command:

| eval early=strftime(early,"%d/m/%Y %H.%M.%S"), late=strftime(late,"%d/m/%Y %H.%M.%S"), StartTime=strftime(StartTime,"%d/m/%Y %H.%M.%S")

Bye.
Giuseppe

Re: Filter records using time modifiers

dsiob — Thu, 27 Jul 2017 09:21:41 GMT

so in place of "-2w@w" I need to use $mytime.Latest$, If I go with eval late=strftime(late,"%d/m/%Y %H.%M.%S") then It does not work If value of $mytime.Latest$ is "now". It is not able to convert "now" to epochtime

Re: Filter records using time modifiers

gcusello — Thu, 27 Jul 2017 10:17:27 GMT

Now() is already in epochtime.
you can see with

index=_internal | head 1 | eval late=now() | table late

beware that in your eval command you losed one %.

Bye.
Giuseppe