Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I've installed Sun Grid Engine on 10 nodes, and one virtual master host.

Now I have to monitor all the resources prior to launching it into production, but I don't know which is the best way. I've tried using xml-qstat, but it seems unstable.

Any tips or suggestions?

Anyone got experience on this?

thanks.

share|improve this question

4 Answers 4

up vote 4 down vote accepted

You could use Ganglia. We use Ganglia with 1000s of nodes at the Holland Computing Center and for the most part, it seems to work fairly well, especially if you are looking for historical graphs. Nagios is used for active monitoring.

share|improve this answer
    
I had gone with this, and it's quite easy and none cpu consuming. Many thanks. –  marc.riera May 10 '10 at 16:14

If I am understanding you correctly you need to monitor bunch of grid servers. What kind of monitoring do you have in mind? Perhaps something like Nagios with some additional scripting could fit your needs?

There is an example over here.

share|improve this answer
    
I thought he meant monetize. –  Ward Apr 21 '10 at 21:46
    
I don't care much about the money. :) And Nagios, it's too generic I think. I'm looking for something which can fit with sun grid engine, maybe hadoop. It's a grid and this means that I must nomitorize each cpu/process separately, not only the service and the performance. –  marc.riera Apr 21 '10 at 22:04
    
Well, Nagios is only as generic as you want it to be. The real value of Nagios is that it can be made to work with anything. So if you figure out how to extract the data you want xml-qstat you can feed it into Nagios. For graphing the data and keeping it for historical purposes you can use Cacti. Again, like with Nagios, you could feed anything into Cacti and get pretty pictures you can show to the management. It's just depends how much effort you want to put into this project. –  solefald Apr 21 '10 at 22:15

Just for the record, also Munin (http://munin-monitoring.org/) is very nice.

share|improve this answer

It sounds like you're more interested in metrics than uptime or availability. Circonus (http://circonus.com/) is a good fit here. You can correlate virtually any metrics, which can be imported over the Resmon XML DTD.

share|improve this answer
    
Looks expensive. –  Tom O'Connor Jan 8 '12 at 16:28

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.