Problem viewing graphs from NDOgrapher

[majed]

read your document but it just mentions what ndo2db does. I remembered that i installed ndo2db just to graph realtime using ndo grapher. i copied it to /var/www and did according to the instructions in

http://exchange.nagios.org/directory/Ad ... er/details

but when i open the graph i get a blank page.
any ideas?

[sreinhardt]

Is there a particular reason you wish to use ndographer over pnp4nagios or something else of that nature? It seems that ndographer has not been updated in over 9 years, this is quite a long time to try and implement with the many changes of core and ndo2db.

[majed]

well latest release states 2009. This ndographer promises realtime graphs. as far as i know pnp4nagios draws graphs minutes old. instead of pnp4nagios i have nagios graph, works fine. Or am i missing something?

[abrist]

Unfortunately, it looks to be a dead project. No information on the website, no documentation. Do realize that this plugin does not offer "realtime" graphing, but "most recent". As he checks get entered into ndo they will be made available in the graph. The only reason why this would be quicker than pnp4nagios would be due to npcd/perfdata processing delays - which really is on average 15seconds to 1 minute.

Unfortunately, I can provide no help for the grapher as the project is technically abandoned.

[majed]

Well thanks for the reply, i guess there is no replacement for cacti then.

[slansing]

I suppose this could technically be a solution:

https://sites.google.com/site/ryanmelis ... -centos-63

Might I ask why you are so adverse to using something such as PNP? I'd personally say getting graph data updated at a sub 15 second -1 minute interval should always come second to a system that is able to show you it's check results. The rate at which graphs are updated as absolutely no bearing on your returned check results being displayed, in fact, that process is split off the moment they return to the Nagios server. The only time that something similar to graphs being updated faster might help would be in a situation such as MRTG where it returns bandwidth checks to RRD files, but even then, those files are checked by a plugin such as check_rrdcached and many times even before being displayed in a graph.

[majed]

pnp is good, i used to use it with nagios 3 on gentoo but it was hard to configure. i tried lately to use nagios graph and it's basically the same but easier to configure.
the nagios cacti plugin seems like a good idea, too bad there isn't a cacti nagios plugin

[abrist]

I am a gentoo guy myself and I never use their ebuilds for core/other nagios stuff as the ebuilds deviate too much from the spec.

Cacti really should not be any faster than pnp, and they definitely do not do "realtime" graphs. Cacti still uses rrds and active checks (snmpgets) with an an interval, so the delay should still be present. Realtime graphs would require a continuous socket connection to an agent. Netflow would be realtime, but snmpgets definitely not.

If you are looking to reduce the delay from a check finishing to the information hitting the rrd, just reduce the interval at which you bulk process perfdata in your nagios.cfg.

[majed]

ok I tried installing cacti but i ran into a problem, perhaps you have an idea?
http://forums.cacti.net/viewtopic.php?f ... 61#p251461

[tmcdonald]

We can't really offer help with installing Cacti as this is the Nagios forum and we already gave our recommendation of pnp4nagios. As abrist said, cacti is not going to provide truly "real-time" graphs as it relies on a rrd file and check intervals. I would suggest keeping with the cacti forum and answering their questions the best you can. More detail is always better.