We upgraded our NetScalers from v9 to v10 so I had to update the NetScaler app to v5. I installed the IPFix v5.0.1 app and it does not work as expected. There are a lot of issues like my host is the heavy forwarder instead of the NetScaler and the source is getting changed to "source = DataSource(address[0], address[1], observer_id)" which appears to be NS_IP:PORT:INSTANCE?
Any way, that's just inconvenient, the real problem is that the app crashes 2, 3, 4 times a day and I have to restart Splunk to get it working again. I have a case open with Splunk. The CRITICAL error that's getting thrown is "UnicodeDecodeError: 'utf8' codec can't decode byte 0x96 in position 967: invalid start byte"; it doesn't like something in our data.
The old v4.8 IPFIX app never had this issue so I was wondering if I could just roll back to that one.
The data is structured differently, so I don't think it will work.
The good news is that it generates the same error every time so I wrote a script that greps the ipfix logs for the error and restarts splunk whenever it occurs. All of our other logs go into syslog so we're only losing 3 minutes of NetScaler logs 3 or 4 times a day.
I assume I can fix the host in props for the NS app.