nagios future?

Michael Friedrich michael.friedrich at univie.ac.at
Thu Jun 11 17:23:15 CEST 2009


Hi there,

Andreas Ericsson wrote:
>
> Merlin has no table prefix. We feel that if you're managing the network
> monitoring server, you have the right to demand a separate database for
> it of the DBA's. table-prefixes are quite handy for things such as forums
> and blogs which are sometimes hosted by home-users on free hosting sites
> where each user gets full access to one database but has no possibility
> to create more. I find it extremely unlikely that Nagios+Merlin will ever
> be installed in an environment where that is the case.

Jep, point taken. I dunno how much Ethan had the experience while coding 
the NDO but it seems the design was done to fit everything which could 
prohibit problems. More or less without regards of the known problems 
right know.
>
> Well, Merlin *is* a rewrite. It doesn't reuse any code at all from 
> NDOUtils
> and it does scale a *lot* better. It replaces NDOUtils for all 
> GUI-related
> activities we do at op5, and the database related parts of the GUI code
> that was previously so arcane it was impossible to debug is now rather
> straight-forward.
And I really like the idea of it - I just need more time to get a deeper 
look and also a running testbox with it. But I am pretty sure I will 
catch on this in a break while developing some dns monitoring tools.
> How can Icinga fit your needs when the underlying model is broken?
> Sorry, but I just don't get it.
Don't get me wrong but we need to apply and change more than 1 single 
Nagios-Server over here. Systems are running fine, no need to change 
that meanwhile. What we need desperately simply is Oracle DB support (it 
works with NDOUtils Oracle but quite unstable).
We've decided to relay on the low ressources we can provide to extend 
the Oracle support by catching up on the most similar system to Nagios. 
While the NDO remains BETA since 2007 and Ethan did only minor bugfixes 
this project is going to die ... and the compatibility will die too if 
Nagios decides to use Merlin instead of NDO - which is a good move - but 
we have to decide:

* update Nagios and everything relaying on the DB Model (we have our own 
tools presenting the data from the DB)
* stay where we are and hope that nothing will fail
* search other solutions where we can contribute better and the 
compatibility will remain

The fork is really good, Ethan is awaken, community is more important 
and development is getting moving again. Let's see what time will bring 
and where we will end :-)
>
> Looks simple enough. Perhaps it'd be worthwhile to test the experimental
> Oracle support in Kohana as well to see if we can get Merlin+Ninja up
> and running fairly quickly. Assuming PHP has Oracle support (which it
> does), I really don't see any issues with that here.
Only compiling instead of package installs, but that shouldn't be that 
big problem for sysadmins.
> I'd prefer if Merlin can use a driver-agnostic layer. Otherwise one would
> have to add support for multiple drivers in Merlin.
>
> There's also the fact that we lack an oracle database to test against,
> so whoever adds oracle support to Merlin and Ninja will have to be
> someone else than me.
As I remarked - I will look up and test Merlin. But meanwhile it should 
be simple to install a testing stage by yourself. For different 
Oracle-Queries you may catch up on the NDOUtils Oracle on nagiosforge.com

Kind regards,
Michael

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null





More information about the Users mailing list