2

I am trying to find plugins or configuration settings to monitor Novell clustered services and was wondering if anyone has done this before or has any info they can share

The system I am trying to monitor runs as two servers accessing configuration files stored on a NAS which in turn presents LUNS using iSCSI over a VLAN. All connections etc work, I am just after getting Nagios to monitor the cluster status, right now we monitor the servers using Nagios with simple ping and standard setting for linux Nagios config so we know if the server is up but we don't know anything regarding the services.

Here is what we currently run for clustered service

Both servers are running Novell OES (Open Enterprise Server) on SLES (Suse Linux Enterprise Server)

Clustered services are (the config files are on the NAS VLAN)

  • iFolder
  • iPrint
  • Netstorage
  • nativehome (shared area for linux desktop home directories)
  • DNS
  • DHCP
  • Number of shared areas for users on windows xp desktop

We have a LAMP server running Nagios Version 3.1.2 which is working a treat for basic connection tests to servers, switches and printer but I would like to make it do more

Hope that all makes sense and sorry if its already been posted elsewhere, I could not find anything for Novell clustered services.

Rodent43
  • 697
  • 3
  • 11

2 Answers2

2

For DNS/DHCP, you can use the existing Nagios plugins from your Nagios server to check that they're DNS-ing and DHCP-ing properly. If your cluster's running on SLES, you should be able to install the Nagios plugins and NRPE on the SLES servers, then monitor that with check_nrpe on the Nagios box. That will let you check mount points, running processes, etc.

RainyRat
  • 3,700
  • 1
  • 23
  • 29
  • Thank you, I will look in to this and post back my progress – Rodent43 Jul 09 '09 at 11:03
  • Thanks RainyRat, I now have the DNS and DHCP monitoring as you suggested, just need to look at NRPE now. – Rodent43 Jul 10 '09 at 08:39
  • Actually, having read geoffc's response, you could probably use check_tcp or check_http to monitor the iFolder/NetStorage from Nagios. – RainyRat Jul 10 '09 at 08:47
  • I tried check_tcp -H -p 80 for iFolder to which I got a response and thought OK lets go...so I then offlined the resource and ran the check again and still got a response...so not sure its a true reflection – Rodent43 Jul 10 '09 at 12:17
  • Probably because the web server's still running, but the app's not - you can specify a URL to use with check_http, so you may want to do that instead. – RainyRat Jul 10 '09 at 14:08
1

I have not yet used NCS on OES Linux, mostly on Netware. There is a series of CLUSTER commands you can execute. Is there a similar one for OES Linux? If so, you could script the check of state with that.

You can look at ports to see if the service is listening.

iFolder, depending, is on :80/443, iPrint is on 631, NetStorage is /NetStorage on 80/443 and you can make a WebDAV connection to NetStorage to test further, so you ought to be able to test that as well.

And so on.

geoffc
  • 2,135
  • 5
  • 25
  • 37
  • I'm not sure on what commands you run on netware, I know you use basic cluster status, cluster view to see which servers are in the cluster and where the services are loaded, is that the commands you mean? – Rodent43 Jul 10 '09 at 08:39
  • Yes, CLUSTER STATUS, CLUSTER RESOURCES both return a formatted list of resources and their current running states. You might also be able to look at the _admin volume, which has a virtual file system that stores in XML the status of the server and some of its services. Definitlty has NSS stats there. – geoffc Jul 10 '09 at 10:04