I set up an alert to trigger a script (sample "echo.bat
" in $SPLUNK_HOME\bin\scripts\
)
The alert fired normal, but the script was not executed.
When I check the splunkd.log, it says:
ERROR ScriptRunner - Couldn't start child process. script="E:\Program Files\Splunk\etc\apps\search\bin\runshellscript.py"
ERROR script - Script execution failed for external search command 'runshellscript.'
I tried to add python.exe path to %PATH
, but it didn't help, strange.....
splunk version: 4.3-115073-x64
windows: server 2008 64-bit
It sounds like it may be the timezone bug that 4.3 has. It was fixed in 4.3.2 and it relates to Splunk firing scripts.
To test do the following;
Navigate to Manager in the top right, select Your Account in the bottom right, change the timezone to, "Default System Timezone". Hit save and try again.
If that works then it is that bug that is affecting you.
Hi Mario, Yes I tried in search bar "xxxx | runshellscript echo.bat arg1 arg2.....", but it reported the runshellscript returned error 1
It sounds like it may be the timezone bug that 4.3 has. It was fixed in 4.3.2 and it relates to Splunk firing scripts.
To test do the following;
Navigate to Manager in the top right, select Your Account in the bottom right, change the timezone to, "Default System Timezone". Hit save and try again.
If that works then it is that bug that is affecting you.
Hi Drainy
I upgraded splunk to 4.3.2 and the bug gone, thank you!
when I set the timezone to default, the splunk page shows "AttributeError: 'str' object has no attribute 'os_startIndex'"
Hi Drainy
If I set the timezone to default, the splunk page will report a common error "str object.....", anyway I will try it, thanks!
did you try to run it manually from the command line to see if you get an error?