We are trying to ingest Peregrine logs for Asset Manager and we can open the log file up on the windows server and it is all ascii text, however when Splunk ingests it, it comes in as \x00[\x00p\x00r\x00a\x00t\x00i and so on. I modified the input from doing a monitor on the log file to powershell:
[powershell://MonitorConnSlots]
script = . "%SPLUNK_HOME%\etc\apps\FWF_scac_logs\bin\MonitorConnSlots.ps1"
interval = 5,35 * * * *
recursive = false
sourcetype - assetmanagerexport
The powershell script has:
cat c:\Program Files\Peregrine\ACFtp\ACExport\MonitorConnectionSlots.log
The first 2 times it ingested, it did it correctly, but then reverted back to the weird x00 stuff afterwards. We have verified with the Peregrine vendor that the logs written out as ASCII text.
Anyone have any ideas?
Looks like it might be a CHARSET issue.
Have you set CHARSET in props.conf for this sourcetype?
Could it be UTF-16?
I've seen a similar problem (though I think the BOM was different) when dealing with Splunk conf files under windows.
I had to force the encoding when opening the files in Python.
...Laurie:{)
Forgot to add, this file only writes out at the top and bottom of every hour incase that makes a difference.