We're running into some issues with the Splunk add-on for Zenoss in our SHC environment, specifically around the credentials that must be stored for the Zenoss environment. Often times, we see the error "sendmodaction - signature="Failed to get password for user xxxx, realm xxxx. Verify credential account exists. User who scheduled alert must have Admin privileges. - 'None' has no attribute 'feed'" in the logs. This seems to happen when...
- A previously saved alert that was working is modified, or...
- Services on the SHC are restarted, or...
- New config bundles are pushed to the SHC from the deployment server.
The add-on (and my configuration of it) works just fine from a single search head (our license master server for example), but that's not really a production solution - it needs to run in the search head cluster. Has anyone been there, done that and have tips? It seems to me like it has something to do with the way the credentials are stored and retrieved within the Add-on.
