Hi @andynewsoncap What about creating a custom app that executes: Linux: date +%Y-%m-%d_%H:%M Windows: Get-Date -format "yyyy-MM-dd_HH:mm" and have your splunk components send the output to your indexers?
... View more
This is now possible to do in SimpleXML.
Modify your input, and add a change action
<default>1</default>
<change>
<set token="display">$label$</set>
</change>
The change your panel title to use the new token
<title>Demand Billed - $display$</title>
... View more
Have you tried to use the $result.fieldname$ token, and change the trigger to "For each result", otherwise you will just get the first row?
... View more
I was sort of able to do this myself. You may want to try
Based on your example above:
| spath input=some.request.complete actions{} | mvexpand actions{} | spath input=actions{} | stats sum("Bar.UPDATED") by "Foo.UPDATED"
... View more
Try this in your props.conf
[your-sourcetype]
KV_MODE = xml
BREAK_ONLY_BEFORE = <ns0:MrmMessages_3_0
NO_BINARY_CHECK = 1
SHOULD_LINEMERGE = true
... View more
I have gotten around this before by using the table command, so in your example :
index= host= |rex= spath |table a,k,h|sort time|append[|dbquery DB "select X,Z,W,P from table_T where Order By time_col"] | table a,k,h,X,Z,W,P
... View more
I had the exact same issue. No matter what I changed the sourcetype to, unless it was "server", which is the default, I got those characters coming through.
I even tried the charset suggestion from here http://answers.splunk.com/answers/24484/sql-server-errorlog, but then on one server I started to get even stranger results.
Only seems to happen with the Cerberus FTP log file though.
... View more
It appears that limit only works for timecharts, but you should be able to get the answer you are wanting by doing a reverse sort, to get the "top" entries to appear first, and then do a head, to show only the first x rows.
So your search string should be :
index=foo | eval length=length(_raw) | chart eval(sum(length)/1024/1024) as MiB by application | sort -MiB | head 15
... View more
I found that changing your timezone, and researching updates the extrapolated time, but I needed to logoff to have the histogram update to the correct time.
... View more
We have recently upgraded to the 5.0.1 release, but are now seeing that we have to bounce splunk daily, otherwise most of the indexes do not get data. Once we bounce the search head and indexer, same instance, then it works for almost a day, and stops again. Any thoughts on trying to diagnose what the issue could be?
... View more