Notifications are disabled

[tonyleatwork]

Another thought is, are there any issues with DNS? Our Unix / Linux systems and Windows systems live on separate domains that cant do reverse lookups of each other.
Also - what are the ramifications of disabling ndo2db?

I found a link topic that looks like my issue but don't have access - can you guys see if this helps? http://www.google.com/url?sa=t&rct=j&q= ... 2529,d.cWc

I am currently engaged in a POC with the intention to purchase, but this is the last thing that is holding us up. Please assist ASAP as my project deadline is coming really close (2 weeks!!)

[tonyleatwork]

After some additional reading and understanding what ndo2db does, I found a file called 'retention.dat' and flushed it. The queue is back down to zero (ipcs -q).

Now it is my understanding that ndo2db is just a means of keeping state data for checks? Does it affect the performance data in anyway? What are the implications of flushing this file?

After flushing the file, the notification checks came back and I can see the hostgroups again.

UPDATE: The issue came back again after a few minutes .. at least I know how to deal with it temporarily...

[abrist]

Have you had an XI quickstart yet? We usually offer it to those evaluating XI. It is a 1 hour remote session with one our techs to help you with any issues you are having during eval.

[tonyleatwork]

Hi -

I have not. That was going to be after our POC and pilot phase when we had enough information to take advantage of the tech. My assigned sales engineer recommended using this forum for support questions pre-purchase.

[slansing]

Yep, this is the right place to be, we are more than happy to help you out. It would be a good idea to take advantage of that quickstart pre-purchase though, if for nothing else but us to come in and verify your system's health.

[tonyleatwork]

Yep, this is the right place to be, we are more than happy to help you out. It would be a good idea to take advantage of that quickstart pre-purchase though, if for nothing else but us to come in and verify your system's health.

Thanks. The quickstart is great but my understanding is that it is a limited resource (one time deal). As far as my environment goes, this VM was built off of an 'official' Nagios XI VM template, so I don't believe that's an issue.

I was able to further narrow down my issue to just the Windows WMI checks (I created a hostgroup with about 6 of them) and once disabled, the kernel message queue frees up and Nagios works fine again. The UNIX group checks (using nrpe) works fine.

If I re-enable the Windows WMI checks group, it will work for about 20 minutes then then queue starts building up again.

Do you guys have access to the enterprise support forum? This post http://www.google.com/url?sa=t&rct=j&q= ... Ww&cad=rja seems to address my specific issue (Nagios checks stop after 20 mins - ndo2db 99% CPU)

[abrist]

The thread mentioned above was resolved in a remote session. It was resolved through the same method posted by Box - by increasing the kernel message limits.
What does load and i/o wait look like on this system? Could you have other resource issues?
Are there any insert/query errors in /var/log/messages?

[tonyleatwork]

Hi -

This is fixed. The culprit was WMI. I guess I had too many WMI service checks that had errors against hosts and check_wmi_plus didn't like that. By adding a -y 2 option and a -t 25 option the issue was resolved.