I'm only using alpha-numeric passwords, so that fix doesn't fit me
I'm more concerned with finding which process launchs that htpasswd with the wrong password. Tomorrow I'll post the audit logs and I'll try to find some kind of pattern because right now this is a complete random issue.
This has been a bit of a "ghost" issue for us as well, so if you find anything out on this, by all means let us know as well. I know that some users have reported that setting all of the security credentials passwords to be the same as their nagiosadmin password permanently fixed their graphs. However, that's not a preferred solution. If there's a process that's changing that backend password, it would make sense why that solution fixed the issue for some people though.
I'll bring up your findings at our staff meeting today, we'll see if we can kick over any more rocks on this as well.
I passed your findings on to our lead developer. This was an issue that we thought we fixed in a previous release, but we may have missed something somewhere. We'll let you know what we find.
It looks like the performance graph code is trying to authenticate to PNP as the "nagiosadmin" user instead of "nagiosxi". The attached script should fix this.
Upload the script to the following directory on your Nagios XI server: