Issues with NSCA

[Carroll, Jim P [Contractor]]

Yes, it does sound like the FIFO is getting a wee bit clogged.

What's your service_reaper_frequency set to?  If it's set to 10, lower it to
5.  Restart nagios and see what happens.

jc

> -----Original Message-----
> From: Ken Snider [mailto:ken.snider at datawire.net]
> Sent: Thursday, January 16, 2003 5:13 PM
> To: nagios-users at lists.sourceforge.net
> Subject: [Nagios-users] Issues with NSCA
> 
> 
> Greetings all,
> 
> I've witnessed a perplexing issue with NSCA, and was hoping 
> someone my have 
> experienced the same, and could perhaps offer a solution to the issue.
> 
> We use NSCA to monitor about 350 passive services onto our 
> network. NSCA is 
> processed through xinetd -reuse, and has Connections per 
> second (cps) and 
> instances set to 200, and UNLIMITED, respectively.
> 
> In essence, xinetd will happily accept up to 200 connections 
> per second.
> 
> We use a wrapper to massage plugin data and pipe to 
> send_nsca, which in turn 
> sends these results to our nagios server. We have about 5 
> plugin results per 
> server, or about 70 servers total doing this, every two minutes.
> 
> What this means is, every two minutes, 70 connections are 
> made to NSCA and 
> processed.
> 
> The *connections* themselves seem to work just fine.. xinetd 
> spawns all the 
> required nsca daemons and they seem to communicate well.
> 
> The problem, however, is that over time, while nagios will 
> first process all 
> these requests without issue, every two minutes, over time, 
> the process 
> seems to take longer and longer, until nagios eventually 
> begins tripping our 
> freshness threshold, which is set at a more-than-generous 500 seconds.
> 
> The box doesn't experience significant load, nor does it seem to be 
> network-related. It appears that, for whatever reason, these 
> requests simply 
> don't make it into the external command pipe, or they are 
> somehow lost in 
> the process.
> 
> I'm hoping someone more familiar with the issue than I can 
> explain *what* 
> NSCA does, when launched from xinetd, and "likely" blocking 
> on the pipe as 
> the requests go into the server.
> 
> I'd hope the solution isn't "less passive checks", or 
> "staggered more", 
> because that suggests that Nagios won't be able to scale as 
> we grow, and 
> we'll reach a point where our passive requests exceed nagios' 
> capability.
> 
> -- 
> Ken Snider
> Senior Systems Administrator
> Datawire Communication Networks Inc.
> 
> 
> 
> -------------------------------------------------------
> This SF.NET email is sponsored by: Thawte.com
> Understand how to protect your customers personal information 
> by implementing
> SSL on your Apache Web Server. Click here to get our FREE 
> Thawte Apache 
> Guide: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0029en
> _______________________________________________
> Nagios-users mailing list
> Nagios-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/nagios-users
> 


-------------------------------------------------------
This SF.NET email is sponsored by: Thawte.com
Understand how to protect your customers personal information by implementing
SSL on your Apache Web Server. Click here to get our FREE Thawte Apache 
Guide: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0029en