Menu
▾
▴
dnx-devel — Developers list
You can subscribe to this list here.
| 2007 |
Jan
|
Feb
(3) |
Mar
(8) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(6) |
Nov
(6) |
Dec
(6) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2008 |
Jan
(14) |
Feb
(3) |
Mar
(10) |
Apr
(73) |
May
(17) |
Jun
(2) |
Jul
(11) |
Aug
(21) |
Sep
(11) |
Oct
(9) |
Nov
(21) |
Dec
|
| 2009 |
Jan
|
Feb
(4) |
Mar
(2) |
Apr
(1) |
May
(7) |
Jun
(6) |
Jul
(1) |
Aug
(5) |
Sep
(1) |
Oct
(1) |
Nov
|
Dec
(1) |
| 2010 |
Jan
(2) |
Feb
(41) |
Mar
(71) |
Apr
(3) |
May
(7) |
Jun
(4) |
Jul
(6) |
Aug
(3) |
Sep
|
Oct
|
Nov
(16) |
Dec
(2) |
| 2011 |
Jan
(4) |
Feb
|
Mar
(8) |
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2012 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Eric S. <esc...@us...> - 2012-02-26 04:02:55
|
Mohammed, Sorry you're having trouble getting DNX working. Unfortunately I don't have a lot of experience with Nagios XI, but I imagine it shares a lot in common with Nagios Core. That being said, I don't believe there will be a 'wizard' available within NagiosXI to help you configure DNX. This will all need to be done from the command line. Furthermore, DNX is designed to offload checks from the main nagios server to achieve scalable load balancing. DNX does not support 'worker node affinity' - there is currently no supported mechanism to distribute only certain checks to DNX workers. Correct me if I'm wrong, but it sounds like you're trying to use a DNX worker to run only certain checks at a remote site. That is not necessarily a good usage case for DNX. If you simply want to distribute existing checks to more physical machines on the same local network, then look no further than DNX. Some other projects exist that might also be worth investigating: http://www.op5.org/community/plugin-inventory/op5-projects/merlin http://labs.consol.de/lang/de/nagios/mod-gearman/ Hope that helps! Eric Schoeller University of Colorado Boulder Office of Information Technology On 02/22/12 02:34, Mohammed Mughal wrote: > This is a message sent by Mohammed Mughal<mm...@bt...> from a > listing named - DNX. The listing URL is: > http://exchange.nagios.org/directory/Addons/Distributed-Monitoring/DNX/details > > Hi, I really hope you can help me with this. I've installed NagiosXI and > configured it with Windows and Linux clients. All is OK. > > I now want to monitor clients in remote networks and off loading the checks > with DNX. > > For my lab I've installed DNX client on Ubuntu 11 and the server on my > NagiosXI server. > > The problem I now have is that how / what configuration wizard do I use to > add a "remote" host/service to my NagiosXI server? > > I've read as much as I can on the net and I'm still not any wiser. > > I'm getting really stressed out trying to get this to work. > > Thanks. > > Mohammed. > > > ------------------------------------------------------------------------------ > Virtualization& Cloud Management Using Capacity Planning > Cloud computing makes use of virtualization - but cloud computing > also focuses on allowing computing to be delivered as a service. > http://www.accelacomm.com/jaw/sfnl/114/51521223/ > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel |
|
From: Mohammed M. <mm...@bt...> - 2012-02-22 09:52:23
|
This is a message sent by Mohammed Mughal <mm...@bt...> from a listing named - DNX. The listing URL is: http://exchange.nagios.org/directory/Addons/Distributed-Monitoring/DNX/details Hi, I really hope you can help me with this. I've installed NagiosXI and configured it with Windows and Linux clients. All is OK. I now want to monitor clients in remote networks and off loading the checks with DNX. For my lab I've installed DNX client on Ubuntu 11 and the server on my NagiosXI server. The problem I now have is that how / what configuration wizard do I use to add a "remote" host/service to my NagiosXI server? I've read as much as I can on the net and I'm still not any wiser. I'm getting really stressed out trying to get this to work. Thanks. Mohammed. |
|
From: Ethan G. <ega...@na...> - 2011-05-09 20:00:25
|
Hi Guys! We'll be having an official Nagios Conference in Minnesota this September and we'd love to have one of the DNX devs give a presentation on using it to achieve greater scalability. Interested? We'd love to have you there and I'm sure many Nagios users would benefit from learning more about DNX and its advantages. More info on the conference is at http://go.nagios.com/nwcna Best regards, Ethan Galstad President ___ Nagios Enterprises, LLC Office: (888)NAGIOS-1 x701 Fax: (651)204-9103 Mobile: (651)278-1477 Email: ega...@na... Web: www.nagios.com |
|
From: Adam A. <aug...@gm...> - 2011-03-28 19:04:53
|
I should apologize for this. I just saw the messages in the queue from John and just approved them without checking. Human error. Sorry for the spam. |
|
From: John C. <joh...@gm...> - 2011-03-28 18:10:59
|
I think it was just a crappy password. I'd been using a trivial dictionary word for years. I've switched my password and security questions since then - my password has numbers in it now, so it's not likely to get cracked with a simple rainbow list. John On 03/28/2011 11:01 AM, Cary Petterborg wrote: > Uh-oh! You got bit. Sorry to give you the bad news. > > Hey, if you figure out how this got done, and how to prevent it, let me know. I've seen this happen to people on Macs as well as windows, and with gmail accounts, not just hotmail, etc. > > Cary > > > > -----Original Message----- > From: John Calcote [mailto:joh...@gm...] > Sent: Thursday, March 24, 2011 5:50 AM > To: nag...@li...; rob...@gm...; dco...@no...; dnx...@li...; an...@or...; dnx...@li...; mar...@us...; ga...@va...; ope...@li...; ban...@gm... > Subject: [Dnx-devel] (no subject) > > http%3A%2F%2Fwww%2Eabesbenchtops%2Ecom%2Eau%2Fimages%2Famw%2Ephp > > ------------------------------------------------------------------------------ > Create and publish websites with WebMatrix > Use the most popular FREE web apps or write code yourself; > WebMatrix provides all the features you need to develop and publish > your website. http://p.sf.net/sfu/ms-webmatrix-sf > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel > > > NOTICE: This email message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. > > > > ------------------------------------------------------------------------------ > Create and publish websites with WebMatrix > Use the most popular FREE web apps or write code yourself; > WebMatrix provides all the features you need to develop and publish > your website. http://p.sf.net/sfu/ms-webmatrix-sf > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel |
|
From: Cary P. <Pet...@ld...> - 2011-03-28 17:01:56
|
Uh-oh! You got bit. Sorry to give you the bad news. Hey, if you figure out how this got done, and how to prevent it, let me know. I've seen this happen to people on Macs as well as windows, and with gmail accounts, not just hotmail, etc. Cary -----Original Message----- From: John Calcote [mailto:joh...@gm...] Sent: Thursday, March 24, 2011 5:50 AM To: nag...@li...; rob...@gm...; dco...@no...; dnx...@li...; an...@or...; dnx...@li...; mar...@us...; ga...@va...; ope...@li...; ban...@gm... Subject: [Dnx-devel] (no subject) http%3A%2F%2Fwww%2Eabesbenchtops%2Ecom%2Eau%2Fimages%2Famw%2Ephp ------------------------------------------------------------------------------ Create and publish websites with WebMatrix Use the most popular FREE web apps or write code yourself; WebMatrix provides all the features you need to develop and publish your website. http://p.sf.net/sfu/ms-webmatrix-sf _______________________________________________ Dnx-devel mailing list Dnx...@li... https://lists.sourceforge.net/lists/listinfo/dnx-devel NOTICE: This email message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. |
|
From: John C. <joh...@gm...> - 2011-03-24 11:50:26
|
http%3A%2F%2Fblog%2Evipsexshop%2Ecom%2Ebr%2Fimages%2Famw%2Ephp |
|
From: John C. <joh...@gm...> - 2011-03-24 11:50:21
|
http%3A%2F%2Fwww%2Eabesbenchtops%2Ecom%2Eau%2Fimages%2Famw%2Ephp |
|
From: William L. <wi...@le...> - 2011-03-21 17:19:05
|
Hahaha. It happened to me just like this a few times although people on this list probably dont know. And I appreciate the early info :). OpenSLP is a good project. On Mon, Mar 21, 2011 at 9:24 AM, John Calcote <joh...@gm...> wrote: > Dnx developers - I apologize. I sent that last note to the wrong list! I > didn't notice it till I was mid-send, and it was too late to hit cancel. > > Sorry, > John > (Also an openslp developer) > > ------------------------------------------------------------------------------ > Colocation vs. Managed Hosting > A question and answer guide to determining the best fit > for your organization - today and in the future. > http://p.sf.net/sfu/internap-sfd2d > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel > |
|
From: John C. <joh...@gm...> - 2011-03-21 16:26:04
|
Dnx developers - I apologize. I sent that last note to the wrong list! I didn't notice it till I was mid-send, and it was too late to hit cancel. Sorry, John (Also an openslp developer) |
|
From: John C. <joh...@gm...> - 2011-03-21 16:25:26
|
Hi all -
If you're monitoring changes to the project, you may have noticed that
the website was changed last night. We've done a complete overhaul.
Please go to:
http://www.openslp.org
and check it out. Let me know if I've really messed anything up. :) Also
let me know of you've got ideas for improvements. h
The new graphics were designed by Alan Adams, one of our project team
members, and I think he did a spectacular job.
Please note that you must use www.openslp.org, not openslp.org. This is
a long-standing DNS problem that has to do with the domain name
purchased for us by Novell. Since Novell owns the domain, I've asked
Alan, a Novell employee, to look into it. (Any word back yet Alan? :) )
John
|
|
From: Adam A. <aug...@gm...> - 2011-01-11 18:00:12
|
DNX is designed currently only distributes service checks. It does not distribute host checks. They are still run on the head. On Tue, Jan 11, 2011 at 2:23 AM, Barbara Martinez <bma...@sa...> wrote: > Hello, > > We have Nagios 3.2.3 running with Distributed Nagios eXecutor but we have a > problem with it and it would be great if you could help us. > All the service checks run remotely but the Host Check is running locally, > which it has no sense because we didn't set any exception in the .cfg file > of the DNX Server. > > ¿Should we configure something we're missing, although the .cfg file > indicates Nagios runs remotely by default? > > Thanks for your help. > Kind regards. > > -- > > Bárbara Martínez Álvarez > Departamento de Sistemas > Centro de Atención al Usuario - SACYL > cau...@sa... > Tlf: 902 555 678 > > --- AVISO SOBRE CONFIDENCIALIDAD. La legislación española ampara el secreto > de las comunicaciones. Este correo electrónico y sus anexos son > estrictamente confidenciales. Si Vd. no es el destinatario del mismo, por > favor, notifíquenoslo inmediatamente y destruya el original. No deberá > copiar este mensaje ni sus anexos o usarlo para propósito alguno, ni > divulgar su contenido a ninguna persona. --- WARNING ABOUT CONFIDENCIALITY. > Spanish law guarantees privacy in electronic communications.This message and > its attachments are confidential. If you are not the intended recipient > please, notify us inmediately and destroy the original. You should not copy > this message or use it for any purpose, nor disclose its content to any > other person. > ------------------------------------------------------------------------------ > Protect Your Site and Customers from Malware Attacks > Learn about various malware tactics and how to avoid them. Understand > malware threats, the impact they can have on your business, and how you > can protect your company and customers by using code signing. > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel > > |
|
From: Barbara M. <bma...@sa...> - 2011-01-11 10:22:05
|
Hello, We have *Nagios 3.2.3* running with Distributed Nagios eXecutor but we have a problem with it and it would be great if you could help us. All the service checks run remotely but the Host Check is running locally, which it has no sense because we didn't set any exception in the .cfg file of the DNX Server. ¿Should we configure something we're missing, although the .cfg file indicates Nagios runs remotely by default? Thanks for your help. Kind regards. -- *Bárbara Martínez Álvarez *Departamento de Sistemas Centro de Atención al Usuario - SACYL cau...@sa... <mailto:cau...@sa...> Tlf: 902 555 678 --- AVISO SOBRE CONFIDENCIALIDAD. La legislación española ampara el secreto de las comunicaciones. Este correo electrónico y sus anexos son estrictamente confidenciales. Si Vd. no es el destinatario del mismo, por favor, notifíquenoslo inmediatamente y destruya el original. No deberá copiar este mensaje ni sus anexos o usarlo para propósito alguno, ni divulgar su contenido a ninguna persona. --- WARNING ABOUT CONFIDENCIALITY. Spanish law guarantees privacy in electronic communications.This message and its attachments are confidential. If you are not the intended recipient please, notify us inmediately and destroy the original. You should not copy this message or use it for any purpose, nor disclose its content to any other person. |
|
From: Eric S. <esc...@us...> - 2011-01-05 23:55:56
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Alberto,<br>
<br>
Nagios only needs to be installed where the DNX Server is running from.
Your DNX Workers (or clients) do not need to have nagios installed. It
sounds like there may be a configuration problem somewhere that is
preventing the workers from requesting jobs. If you're certain both the
worker and server can see each other on the network, you also may want
to check the firewalls on both hosts. Then, if you're certain there is
no network or firewall problem, you can copy snippets of the debug logs
to the list and we can try to help diagnose the issue.<br>
<br>
Thanks,<br>
<br>
Eric<br>
<br>
Alberto García Pérez wrote:
<blockquote cite="mid:BLU...@ph..."
type="cite">
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style> <br>
Hi,<br>
<br>
I want to deploy DNX.<br>
I have installed DNX Server in a Nagios Central Server Machine and DNX
Client in another machine.<br>
The machines have communication and the event broker module is running
succesfully.<br>
<br>
Is necessary to installed Nagios in the Client machine?<br>
The client machine is waiting and the server too but they don't receive
checks.<br>
<br>
Sorry for my english.<br>
Kind regards.<br>
<br>
<pre wrap="">
<hr size="4" width="90%">
------------------------------------------------------------------------------
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and,
should the need arise, upgrade to a full multi-node Oracle RAC database
without downtime or disruption
<a class="moz-txt-link-freetext" href="http://p.sf.net/sfu/oracle-sfdevnl">http://p.sf.net/sfu/oracle-sfdevnl</a></pre>
<pre wrap="">
<hr size="4" width="90%">
_______________________________________________
Dnx-devel mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Dnx...@li...">Dnx...@li...</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/dnx-devel">https://lists.sourceforge.net/lists/listinfo/dnx-devel</a>
</pre>
</blockquote>
</body>
</html>
|
|
From: Alberto G. P. <alb...@ho...> - 2011-01-03 20:13:55
|
Hi, I want to deploy DNX. I have installed DNX Server in a Nagios Central Server Machine and DNX Client in another machine. The machines have communication and the event broker module is running succesfully. Is necessary to installed Nagios in the Client machine? The client machine is waiting and the server too but they don't receive checks. Sorry for my english. Kind regards. |
|
From: Eric S. <esc...@us...> - 2010-12-01 16:33:33
|
Assaf, I think we may want to have specific mod_gearman discussions on that project's mail list. https://groups.google.com/group/mod_gearman (even though I'd prefer a standard mailman list over google anyday!!) Eric Assaf Flatto wrote: > Hello All > > I am trying to test the mod_gearman implementation with nagios 3.2.3 on > a 64bit SLES 10.3, i've build the server , with all the complications > that came with it (having to install gearman 0.14 _ as the mod_gearman > 0.8 - will not run on any older versions) , having to use libevent 1.4 > since the mod_gearman wouldn't work with libevent 1.1 ( OS supplied) . > > The main issue i have is that we are using gearman on our platform to > process tasks , but doing a libevent upgrade and greaman upgrade on all > servers is not an option . > > I have several question : > > 1) will the greaman server (nagios , gearman 0.14) work with older > greaman workers ( 0.12) ? > > if it is possible then > > 2) how to set up a gearman worker on a sepreate machine - the > documentation seems to focus on the same version only - and not take in > to account older version setup. > > Thanks > > |
|
From: Assaf F. <na...@fl...> - 2010-12-01 11:29:27
|
Hello All I am trying to test the mod_gearman implementation with nagios 3.2.3 on a 64bit SLES 10.3, i've build the server , with all the complications that came with it (having to install gearman 0.14 _ as the mod_gearman 0.8 - will not run on any older versions) , having to use libevent 1.4 since the mod_gearman wouldn't work with libevent 1.1 ( OS supplied) . The main issue i have is that we are using gearman on our platform to process tasks , but doing a libevent upgrade and greaman upgrade on all servers is not an option . I have several question : 1) will the greaman server (nagios , gearman 0.14) work with older greaman workers ( 0.12) ? if it is possible then 2) how to set up a gearman worker on a sepreate machine - the documentation seems to focus on the same version only - and not take in to account older version setup. Thanks -- Never,Ever Cut A Deal With a Dragon Next year I will be doing the London to Paris bike ride to raise money for the DogTrust (www.dogstrust.co.uk) . Please Sponsor me at http://www.justgiving.com/Assaf-Flatto |
|
From: Sven N. <sv...@ni...> - 2010-11-18 11:05:58
|
On 11/17/10 18:22, Adam Augustine wrote: > I imagine that with > some simple patches to Nagios it could also do notifications as well. I tried that, but as you said, it won't be possible without changes to the nagios core. Then i realized that its probably not a good idea to distribute notifications at all. At least you would have to distinguish between normal notifications and notifications for the distributed infrastructure. Most notification commands should be really fast and do hopefully not run very often :-) But if so, you still could use the send_gearman command or anything similar to create custom jobs. And i saw nagios configurations which use eventhandler to connect a ticketing system. This would be possible already. Regards, Sven |
|
From: Jason B. <ja...@fo...> - 2010-11-18 09:16:33
|
Forget about ZeroMQ :-) What I was talking about was an embeddable queueing library which is what Gearman actually is. I hadn't looked at the code well enough when I made the suggestion and assumed Gearman was an external service. Gearman looks perfect for most people's needs I think. Jason > Hi, > > On 11/18/10 2:29, Jason Benguerel wrote: >> Honestly I've had time to look at the code and I think that Gearman is a better solution than ZeroMQ. > > The good and bad thing about Gearman is its simplicity. This makes gearman really fast but has a few downsides. For example there is no way of access control. > If there is a better free job queueing system somewhere, it should be easy to change or to support both. As you already said, gearman is only the transport layer. > > Regards, > Sven > > ------------------------------------------------------------------------------ > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today > http://p.sf.net/sfu/msIE9-sfdev2dev > _______________________________________________ > Dnx-users mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-users |
|
From: Sven N. <sv...@ni...> - 2010-11-18 09:12:04
|
Hi Adam,
On 11/17/10 18:21, Adam Augustine wrote:
> I just discovered the mod_gearman project recently as well. I haven't
> actually tested it yet, but my quick review is that it is mostly what
> we were trying to do back when we started DNX.
As i said, i am following the dnx project for years now, so i tried to solve all the things i noticed.
> Anyway, mod_gearman actually looks really good (again, just my brief
> look at it so far)
Thanks :-)
> implement the functionality into DNX. I have seen Sven's name on
> several projects we use (WebInject and others, thanks for picking that
> up by the way) and all indications are he knows what he is doing.
Youre welcome. In fact i did a complete rewrite of webinject. But its compatible with old configs.
>> (I will port to AMQP when the protocol gets stabilized)
I had a look at AMQP but it didn't work for me, so i switched to gearman and i am happy so far.
The only thing is the missing authentication.
> That is how Node Affinity should be implemented (IMNSHO), via the new
> (to Nagios 3.x anyway) custom object variables, say
> _dnx_node_affinity=<name of cluster of worker nodes>, where <name of
> cluster of worker nodes> is defined in the DNX config files (possibly
> being modifiable on the fly).
The reason for choosing host/servicegroups is, that they already exist
and most people have grouped their hosts and services already in useful ways.
So you could just load the plugin and use the existing groups.
What i usually create is a meta hostgroup:
define hostgroup {
hostgroup_name worker_emea
hostgroup_members group1,group2,group3
}
so i have a single place to seperate groups for my worker.
Next thing is, you can display these groups in your webgui which is not possible for custom variables.
>> Now messaging middleware isn't something that everyone wants to run, and it adds a whole new layer to have to maintain (and debug when things go wrong)
Thats true. Luckily the gearman job server is really stable. I hadn't a single
problem with it so far. But right, you definitly have to add some monitoring for it.
The plugins for monitoring gearman infrastructure is included.
> I don't know how difficult Gearman is to set up, but it seems on the
> surface at least to resemble the DNX design with a client (the NEB
> module), Job Server (the dnxServer daemon), and workers. I am looking
> over the tutorial... it seems you need to specify each command you
> want executed in advance... but I imagine mod_gearman addresses
> that...
I don't understand. The only thing you have to do after setting everything up is to make sure your plugins stay up to date. DNX offers a rsync solution for that. There is currently no solution for mod_gearman, but i expect everyone how is setting up distributed monitoring has already a solution for software distribution or is familiar with rsync.
And sorry, i didn't want to hijack your mailing list for mod_gearman.
Regards,
Sven
|
|
From: Sven N. <sv...@ni...> - 2010-11-18 08:48:05
|
Hi, On 11/18/10 2:29, Jason Benguerel wrote: > Honestly I've had time to look at the code and I think that Gearman is a better solution than ZeroMQ. The good and bad thing about Gearman is its simplicity. This makes gearman really fast but has a few downsides. For example there is no way of access control. If there is a better free job queueing system somewhere, it should be easy to change or to support both. As you already said, gearman is only the transport layer. Regards, Sven |
|
From: Jason B. <ja...@fo...> - 2010-11-18 01:29:25
|
Honestly I've had time to look at the code and I think that Gearman is a better solution than ZeroMQ. This project is extremely well written and can do everything DNX does but adds affinity and a better transport layer. It is possible that DNX could outperform this because of UDP and the simpler ring buffers, but both of those things sacrifice reliability and really aren't worth any performance gain IMHO. I would say that they took the great ideas of DNX, combined them with the concept of a true message queue and added an affinity layer. A trivial patch would be required to use custom Nagios object attributes to allow a more fine grained affinity that some require. Only the fact that I have based our architecture around a separate message queue prevents me from using this, and the embedded nature of the gearman libs makes swappable messaging architectures a no starter, but for anyone who hasn't invested in a messaging queue this solution looks outstanding. Given the limited development support available for DNX I'd recommend strongly that DNX users evaluate this software and DNX developers consider supporting this codebase as an alternative to further development. I think DNX is an outstanding project, and I learned a great deal about C and Nagios internals from this codebase, but having adding affinity to it using my best attempt at maintaining it's internal concepts, I feel it is at a dead end as far as being able to properly add these kinds of abilities. But that's just my take on it, and it is certainly not meant to insult the developers or try to wish for it's demise. Jason On Nov 18, 2010, at 2:22 AM, Adam Augustine wrote: > On Mon, Nov 15, 2010 at 2:18 PM, Sven Nierlein <sv...@ni...> wrote: >> On 11/15/10 21:36, William Leibzon wrote: >>> DNX is maintained. >> >> Thats good to hear. However, i asked the question about putting hostgroup affinity into the core in march. But my questions got never answered. I am watching the DNX project over several years now, always looking forward for this feature. >> >> Btw, the hostgroup affinity patch is located here: https://github.com/Bakafish/DNX_Affinity/ >> >> I gave up waiting and wrote mod_gearman. mod_gearman tries a different approach. It uses already developed opensource software where possible and thus using the Gearman Jobserver for loadbalancing. There is hostgroup/servicegroup affinity already included in the core. And it is possible to distribute host checks and eventhandler too. >> If you want to have a look: https://github.com/sni/mod_gearman >> >> And to be complete, you should also have a look at the merlin project, which also has hostgroup affinity included. >> >> Regards, >> Sven >> > > That is my fault, and I apologize. I should have been more responsive. > We were in the middle of some major problems here at work and the list > was consumed with bugs related to that. It didn't penetrate my thick > skull that you were offering patches (thanks for the several bug fixes > you did manage to get through to us by the way). Had I known I would > have added you as a developer. > > I can't speak for Bob and John or the others who are the major > contributors, but I know we don't have a lot in the way of resources > to spend on DNX development. It may be that we will get those > resources in the next few months, but I certainly have no guarantees. > I don't want to get into the situation that Nagios is currently in, > where developmental progress still depends on a single person. I am > willing to open up commit and release rights to those who have > demonstrated competence, even if I don't necessarily completely agree > with their direction, as long as there is rough consensus and running > code. > > I am also open to the discussion of throwing our weight behind > mod_gearman as a potential replacement for DNX. It looks like it might > need a few additional features to be as transparent as DNX tries to > be, but I doubt that would be difficult to get implemented, depending > on peoples time of course. > > It already has features we wanted, like node affinity, hostchecks, > eventhandlers, and (for us, a big one) encryption. I imagine that with > some simple patches to Nagios it could also do notifications as well. > I don't know how hard it would be to make it use custom attributes to > handle node affinity. > > The current DNX is fairly robust to worker node failure, > re-distributing load automatically to compensate. It also, by nature > of the algorithm, should load balance according to a workers ability, > meaning a less capable machine will attempt to execute less jobs than > a more capable machine. It would be good to see how mod_gearman > handles those issues. > > Community, what are your thoughts? > > ------------------------------------------------------------------------------ > Beautiful is writing same markup. Internet Explorer 9 supports > standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. > Spend less time writing and rewriting code and more time creating great > experiences on the web. Be a part of the beta today > http://p.sf.net/sfu/msIE9-sfdev2dev > _______________________________________________ > Dnx-users mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-users |
|
From: Adam A. <aug...@gm...> - 2010-11-17 17:22:24
|
On Mon, Nov 15, 2010 at 2:18 PM, Sven Nierlein <sv...@ni...> wrote: > On 11/15/10 21:36, William Leibzon wrote: >> DNX is maintained. > > Thats good to hear. However, i asked the question about putting hostgroup affinity into the core in march. But my questions got never answered. I am watching the DNX project over several years now, always looking forward for this feature. > > Btw, the hostgroup affinity patch is located here: https://github.com/Bakafish/DNX_Affinity/ > > I gave up waiting and wrote mod_gearman. mod_gearman tries a different approach. It uses already developed opensource software where possible and thus using the Gearman Jobserver for loadbalancing. There is hostgroup/servicegroup affinity already included in the core. And it is possible to distribute host checks and eventhandler too. > If you want to have a look: https://github.com/sni/mod_gearman > > And to be complete, you should also have a look at the merlin project, which also has hostgroup affinity included. > > Regards, > Sven > That is my fault, and I apologize. I should have been more responsive. We were in the middle of some major problems here at work and the list was consumed with bugs related to that. It didn't penetrate my thick skull that you were offering patches (thanks for the several bug fixes you did manage to get through to us by the way). Had I known I would have added you as a developer. I can't speak for Bob and John or the others who are the major contributors, but I know we don't have a lot in the way of resources to spend on DNX development. It may be that we will get those resources in the next few months, but I certainly have no guarantees. I don't want to get into the situation that Nagios is currently in, where developmental progress still depends on a single person. I am willing to open up commit and release rights to those who have demonstrated competence, even if I don't necessarily completely agree with their direction, as long as there is rough consensus and running code. I am also open to the discussion of throwing our weight behind mod_gearman as a potential replacement for DNX. It looks like it might need a few additional features to be as transparent as DNX tries to be, but I doubt that would be difficult to get implemented, depending on peoples time of course. It already has features we wanted, like node affinity, hostchecks, eventhandlers, and (for us, a big one) encryption. I imagine that with some simple patches to Nagios it could also do notifications as well. I don't know how hard it would be to make it use custom attributes to handle node affinity. The current DNX is fairly robust to worker node failure, re-distributing load automatically to compensate. It also, by nature of the algorithm, should load balance according to a workers ability, meaning a less capable machine will attempt to execute less jobs than a more capable machine. It would be good to see how mod_gearman handles those issues. Community, what are your thoughts? |
|
From: Adam A. <aug...@gm...> - 2010-11-17 17:21:28
|
On Mon, Nov 15, 2010 at 6:10 PM, Jason Benguerel <ja...@fo...> wrote: > Taking a quick glance at the mod_gearman project, it's really the correct sort of approach in my opinion. My experimental branch of DNX taught me that attempting to emulate a queuing system using ring buffers was a poor fit, and trying to improve the robustness of the UDP transport also was sub optimal. Once I started researching message queues I realized that this was a much better solution for the whole dispatch and retrieval layer implemented in DNX for WAN systems. Although I do think that if you are only running DNX clients within your LAN that the UDP non-affinity approach is a much more lightweight solution. > I just discovered the mod_gearman project recently as well. I haven't actually tested it yet, but my quick review is that it is mostly what we were trying to do back when we started DNX. When we went down the UDP route, most projects for queuing didn't exist, weren't mature enough, or we just couldn't find them. That is why we implemented our own (risking not doing it "right"). Our very first attempt at transparent clustering was using the *Mosix* clustering solutions, but that was dying quickly. Bob was able to code something up relatively quickly and it worked pretty well, so we pressed forward. When this was started Nagios 2.x was still pre-beta and there really weren't a lot of options (that we could find at least). Anyway, mod_gearman actually looks really good (again, just my brief look at it so far), and I would be happy to accept patches to implement the functionality into DNX. I have seen Sven's name on several projects we use (WebInject and others, thanks for picking that up by the way) and all indications are he knows what he is doing. > The limitations of trying to reinvent a message queue is why I stopped work on that fork (which may have given a false impression that DNX itself wasn't being worked on) in favor of something much like this mod_gearman project. The only difference that I can see to what I've been working on is that I tried to make my solution compatible with several queueing solutions by implementing it with the STOMP protocol (I will port to AMQP when the protocol gets stabilized) and leveraging the custom attributes in the Nagios configuration to allow for more flexible affinity (in addition to simple Hostgroup style affinity.) > That is how Node Affinity should be implemented (IMNSHO), via the new (to Nagios 3.x anyway) custom object variables, say _dnx_node_affinity=<name of cluster of worker nodes>, where <name of cluster of worker nodes> is defined in the DNX config files (possibly being modifiable on the fly). Logically, you should be able to have a single custom object variable on a service, host, servicegroup, or hostgroup, and have it do the logical thing (that check or group of checks would execute on that named cluster of one or more worker nodes). Precedence would have to be worked out (a variable on the service should override the variable on the host, etc). > Now messaging middleware isn't something that everyone wants to run, and it adds a whole new layer to have to maintain (and debug when things go wrong), and so it's probably not the ideal approach for the DNX project which I think fills a niche where the user can compile the NEB and just run it without any significant configuration changes (which was why I used hostgroups for affinity in the first place.) The ZeroMQ project seemed like a nice embeddable solution, and maybe DNX could leverage that as there wouldn't be any need to have to set up a standalone queue. The custom object attributes are a lot easier to access programmatically than I had realized and should be leveraged as well. But the current situation is what it is, with a UDP transport and no affinity DNX is severely limited for true enterprise level monitoring, and should either target the limited scope of LAN usage or come up with a viable solution to these two issues. > > Jason > Because we knew that the UDP method wasn't ideal, DNX was designed to have the option of specifying different transport types. Some of that cleanness of code separation may have rotted a bit over time, but could probably be restored if it isn't completely working today. We lost the full time developer position we had and the we haven't been able to push DNX development like we would have liked. But I think we can still review patches. I don't know how difficult Gearman is to set up, but it seems on the surface at least to resemble the DNX design with a client (the NEB module), Job Server (the dnxServer daemon), and workers. I am looking over the tutorial... it seems you need to specify each command you want executed in advance... but I imagine mod_gearman addresses that... Having DNX support both ZeroMQ and Gearman would be extremely nice. Anyone offering? Seriously, I am willing to open up development quite a bit now (see my reply to Sven). |
|
From: Jason B. <ja...@fo...> - 2010-11-16 01:39:19
|
Taking a quick glance at the mod_gearman project, it's really the correct sort of approach in my opinion. My experimental branch of DNX taught me that attempting to emulate a queuing system using ring buffers was a poor fit, and trying to improve the robustness of the UDP transport also was sub optimal. Once I started researching message queues I realized that this was a much better solution for the whole dispatch and retrieval layer implemented in DNX for WAN systems. Although I do think that if you are only running DNX clients within your LAN that the UDP non-affinity approach is a much more lightweight solution. The limitations of trying to reinvent a message queue is why I stopped work on that fork (which may have given a false impression that DNX itself wasn't being worked on) in favor of something much like this mod_gearman project. The only difference that I can see to what I've been working on is that I tried to make my solution compatible with several queueing solutions by implementing it with the STOMP protocol (I will port to AMQP when the protocol gets stabilized) and leveraging the custom attributes in the Nagios configuration to allow for more flexible affinity (in addition to simple Hostgroup style affinity.) Now messaging middleware isn't something that everyone wants to run, and it adds a whole new layer to have to maintain (and debug when things go wrong), and so it's probably not the ideal approach for the DNX project which I think fills a niche where the user can compile the NEB and just run it without any significant configuration changes (which was why I used hostgroups for affinity in the first place.) The ZeroMQ project seemed like a nice embeddable solution, and maybe DNX could leverage that as there wouldn't be any need to have to set up a standalone queue. The custom object attributes are a lot easier to access programmatically than I had realized and should be leveraged as well. But the current situation is what it is, with a UDP transport and no affinity DNX is severely limited for true enterprise level monitoring, and should either target the limited scope of LAN usage or come up with a viable solution to these two issues. Jason On Nov 16, 2010, at 6:18 AM, Sven Nierlein wrote: > On 11/15/10 21:36, William Leibzon wrote: >> DNX is maintained. > > Thats good to hear. However, i asked the question about putting hostgroup affinity into the core in march. But my questions got never answered. I am watching the DNX project over several years now, always looking forward for this feature. > > Btw, the hostgroup affinity patch is located here: https://github.com/Bakafish/DNX_Affinity/ > > I gave up waiting and wrote mod_gearman. mod_gearman tries a different approach. It uses already developed opensource software where possible and thus using the Gearman Jobserver for loadbalancing. There is hostgroup/servicegroup affinity already included in the core. And it is possible to distribute host checks and eventhandler too. > If you want to have a look: https://github.com/sni/mod_gearman > > And to be complete, you should also have a look at the merlin project, which also has hostgroup affinity included. > > Regards, > Sven > > ------------------------------------------------------------------------------ > Centralized Desktop Delivery: Dell and VMware Reference Architecture > Simplifying enterprise desktop deployment and management using > Dell EqualLogic storage and VMware View: A highly scalable, end-to-end > client virtualization framework. Read more! > http://p.sf.net/sfu/dell-eql-dev2dev > _______________________________________________ > Dnx-devel mailing list > Dnx...@li... > https://lists.sourceforge.net/lists/listinfo/dnx-devel |
4 messages has been excluded from this view by a project administrator.
