search command
host= index= sourcetype=syslog job=* "jobname" | dedub job | fields - _raw | timechart span=1d count by jobname
I get the result as
_time jobname
2018-09-08 24
2018-09-07 12
2018-09-06 36
But I need the result as below ( like dividing the jobname field value by 6)
_time jobname
2018-09-08 4
2018-09-07 2
2018-09-06 6
Please suggest.
Maybe the shortest solution is to write "dedup" (instead of dedub)?
it is already dedup.. its a typo here.
You are searching for job=* "jobname"
, you dedup by job
and timechart by jobname
. In another post you have name1
and url
(the latter as a field name). So what are you really searching for?
And regarding this "devided by 6" - do you really want to divide? Why are you sure your number should be "like dividing the jobname by 6"?
Would be great to get an understanding of the underlying data.
name1 is the actual.. I had explained with example. I need not to be exactly 6 it can be any number but I want to specify.
@renjith.nair nothing worked
the log contains job ids and urls. URL ex: http://website/names
the query is
host="hostname" index="indexname" sourcetype="syslog" job=* "http://website/name1" | dedup job | fields - _raw
| timechart span=1d count by url
I get the result as
_time jobname
2018-09-08 24
2018-09-07 12
2018-09-06 36
But I need the result as below ( like dividing the jobname field value by 6)
_time jobname
2018-09-08 4
2018-09-07 2
2018-09-06 6
Both of the solutions provided by @renjith.nair should have worked. Here's a third...
host=foo index=bar sourcetype=syslog job=* "jobname"
| dedup job
| eval quarter= 0.25
| timechart span=1d sum(quarter) by url
And here's a fourth ...
host=foo index=bar sourcetype=syslog job=* "jobname"
| dedup job
| timechart span=6h count by url
| bin _time span=1d
| stats avg(*) as * by _time
And I'm tempted to come up with a fifth, just for grins.
All of these should work, given your stated requirements.
hi @mindia,
Did the answer below solve your problem? If so, please resolve this post by approving it!
If your problem is still not solved, keep us updated so that someone else can help ya.
Thanks for posting!
Not yet.....
Logs contain job ids and urls
Ex: job=23 url=http://website/names
Make a note, url contains names, whereas I am searching for each name. like name1, name2, etc.
host= index= sourcetype=syslog job=* "http://website/name1" | dedub job | fields - _raw | timechart span=1d count by url
Splunk search returns results like below
_time name1
2018-09-08 24
2018-09-07 12
2018-09-06 36
But I need the result as below ( like dividing the name1 field value by 6)
_time name1
2018-09-08 4
2018-09-07 2
2018-09-06 6
@mindia,
There are multiple ways to do it.
host= index= sourcetype=syslog job=* "jobname" | dedub job | fields - _raw | timechart span=1d count by jobname|untable _time,jobname,count|eval count=count/6|xyseries _time,jobname,count
OR
host= index= sourcetype=syslog job=* "jobname" | dedub job | fields - _raw | timechart span=1d count by jobname|foreach * [eval <<FIELD>>=<<FIELD>>/6]
Actually both commands did not help here. My search is slightly different.
The logs contain job ids and url = "https://url/
I am searching for the and count by I am specifying url
host= index= sourcetype=syslog job=* "" | dedub job | fields - _raw | timechart span=1d count by url
I tried below search using foreach
host= index= sourcetype=syslog job=* "" | dedub job | fields - _raw | timechart span=1d count by url |foreach * [eval <>=<>/6]
.. it displayed the same result.. actually what to substitute for FIELD is not clear.
The another commend using untable was not able execute. it gave errors.
The logs contain job ids and url = "https://url/name1
I am searching for the name1 and count by I am specifying 'url'
host= index= sourcetype=syslog job=* "name1" | dedub job | fields - _raw | timechart span=1d count by url
similarly I would try for name2
host= index= sourcetype=syslog job=* "name2" | dedub job | fields - _raw | timechart span=1d count by url
@mindia, for the foreach you dont need to substitute anything , just use FIELD as it is.
What's the error in untable ?
In the result after timechart , how many columns are you getting ? I mean is it _time,jobname1,jobname2 or is it only one job name ?
Below is a run anywhere example you could use and compare
index=_* earliest=-15m |timechart span=5m count by sourcetype|untable _time,sourcetype,count|eval count=count/6|xyseries _time,sourcetype,count