Nagios Support Forum

Hi,

I don't understand the following behaviour with every Nagios XI reload. With every reload the nsca instances grow.

The problem with this is that if you don't keep an eye it is possible to get a nsca limit and there would be a problem with passive alerts. I had this problem a few times and don't want to set the xinetd.conf instances value to UNLIMITED.

With a Nagios reload from the command line this doesn't happen.

Is it a configuration problem?

Greetings,
Jorge.

I'm not sure I quite understand. Can you provide some more details - exact steps, screenshots, output, etc... ? What version of XI are you using? How many hosts do you have sending data via nsca? How frequently is data getting sent?

It has nothing to do with how many hosts are sending data trough NSCA.

1. Begin with a clean scenario.

1. No nagios process running or processes with the user nagios (nagios, ndo2db, npcd, cron jobs...).
2. shutdown xinetd.

3. Start xinetd
4 Start ndo2db
5. Start nagios
6. Start npc
7. ...

Now you can check with the comand: ps aux | grep nsca one or a few nsca instances. And the systems keeps stable it handles well the incomming data from NSCA checks as long as you don't apply a new configuration in Nagios XI.

Without applying a new configuration in Nagios XI the nsca instances don't grow.

If you apply a new configuration in Nagios XI. You can check again with: ps aux | grep nsca that there are more instances.

Once you reached the configuration variables INSTANCES threshold in xinetd.conf you won't get inbound data from NSCA because it reaches it's limit (you can see that the nsca fails because it reaches its limit in /var/log/messages or other configured log). For example if you have set the configuration value to 50 and your ps aux shows 50 nsca instances. I'ts unnecesary to raise the configuration value because every time you apply a Nagios XI configuration the instances grow. Raising the value higher or set it to UNLIMITED coud compromise the security of the system and you could reach a maximum of open files, and getting out of memory.

If nsca is running under xinetd it will not show up when running Additionally, applying configuration has no affect on NSCA unless someone modified some other init script.

What OS and XI version is this?

Can you show the results of the following