Yes, your mileage may vary. Sorry, bad to use pointless acronyms.mkmurray wrote:YMMV = Your mileage may vary?
I had to Google it, I've never seen that acronym before.
Distributed Nagios eXecutor (DNX)
-
- New Member
- Posts: 3
- Joined: Wed Mar 14, 2007 8:46 am
-
- New Member
- Posts: 4
- Joined: Mon May 14, 2007 10:07 pm
- Location: Littleton, Crawlorado
- Contact:
Feedback...
I am:
1) New to this forum....
2) Not completely familiar with the environment being monitored.
I do have one question though... Presuming you're monitoring a Unix environment, wouldn't the NRPE plugin provide similar functionality?
If you're monitoring Windows, then ignore my post. The Windows monitor plugin does work but is not the most graceful implementation I've seen. I really like the experience I have with NRPE and highly recommend it for anyone monitoring Linux/Unix systems.
Regards,
Robert
1) New to this forum....
2) Not completely familiar with the environment being monitored.
I do have one question though... Presuming you're monitoring a Unix environment, wouldn't the NRPE plugin provide similar functionality?
If you're monitoring Windows, then ignore my post. The Windows monitor plugin does work but is not the most graceful implementation I've seen. I really like the experience I have with NRPE and highly recommend it for anyone monitoring Linux/Unix systems.
Regards,
Robert
-
- New Member
- Posts: 1
- Joined: Sat Jul 07, 2007 12:10 am
DNX development
First of all, thank you very much or this nice contribution to the community.
With alot of interest I checked out the DNX package.
I'm wondering if there still is development on this package?
For my monitoring setup, the dnx is not suitable (yet).
We monitor worldwide different private networks, for which we setup different checks.
It would be really usefull, if you can specify which workers can deal with which hosts. In this case, we can still use dnx, with the flexibility of monitoring geographical seperated networks.
This would be a more than ideal setup for us, with great scalability.
Hope you can have some feedback for me.
With alot of interest I checked out the DNX package.
I'm wondering if there still is development on this package?
For my monitoring setup, the dnx is not suitable (yet).
We monitor worldwide different private networks, for which we setup different checks.
It would be really usefull, if you can specify which workers can deal with which hosts. In this case, we can still use dnx, with the flexibility of monitoring geographical seperated networks.
This would be a more than ideal setup for us, with great scalability.
Hope you can have some feedback for me.
-
- New Member
- Posts: 1
- Joined: Fri Jul 13, 2007 8:14 am
DNX - feedback and questions
Hi
I do some tests with DNX; Results :
DNX have some problems with 30000 services or more,
but DNX woks well with 20000 services and 1500 hosts,
BUT i have about 50 services with state : DNX Service Check Timed Out
How I can correct it ?
I have an other problem with clients. Clients stop themself...
Syslog says :
How I can correct it ?
Titilambert
Me too... please answer usjoop3 wrote: I'm wondering if there still is development on this package?
It would be over cool ...joop3 wrote: It would be really usefull, if you can specify which workers can deal with which hosts. In this case, we can still use dnx, with the flexibility of monitoring geographical seperated networks.
I do some tests with DNX; Results :
DNX have some problems with 30000 services or more,
but DNX woks well with 20000 services and 1500 hosts,
BUT i have about 50 services with state : DNX Service Check Timed Out
How I can correct it ?
I have an other problem with clients. Clients stop themself...
Syslog says :
Code: Select all
/var/log/syslog.0:Jul 12 21:27:53 dnx4-desktop dnxClient[10992]: initWorkerComm: dnxChanMapAdd(Dispatch) failed for thread 3aa46b90: 2
/var/log/syslog.0:Jul 12 21:27:53 dnx4-desktop dnxClient[10992]: dnxWorker[3aa46b90]: Failed to initialize thread comm channels: 2
Titilambert
- mkmurray
- Senior Member
- Posts: 3266
- Joined: Tue Jan 23, 2007 9:56 pm
- Location: Utah
- Contact:
Perhaps technical questions about DNX would be better suited on its SourceForge page. Thanks.titilambert wrote:I have an other problem with clients. Clients stop themself...
Syslog says :How I can correct it ?Code: Select all
/var/log/syslog.0:Jul 12 21:27:53 dnx4-desktop dnxClient[10992]: initWorkerComm: dnxChanMapAdd(Dispatch) failed for thread 3aa46b90: 2 /var/log/syslog.0:Jul 12 21:27:53 dnx4-desktop dnxClient[10992]: dnxWorker[3aa46b90]: Failed to initialize thread comm channels: 2
-
- New Member
- Posts: 1
- Joined: Sun Dec 09, 2007 4:29 pm
- Contact:
Thank you for the information on Nagios, I mentioned it to my co-workers and they said they wanted it setup and working 5 years ago. So I installed it on a ubuntu server machine, and have it linked to 30 host and 170 services so far. Those cfg files can get long fast so I understand the reason behind only entering it all in once. I'm looking at ways to monitor remote networks and well as our internal network. We have clients that are embedded in larger hospital networks, and would like to have the ability to somehow route traffic through one port and check server stats from Nagios. We do have other less intrusive systems that we use such as Bomgar, whcih not only allows us to monitor basic system settings, but allows us to use jumppoints to access the machines remotely and even support our clients in real time while talking to them on the phone in our office.
I saw a nice little addon that monitors window clients via WMI. You only have to install the agent on an IIS webserver, and you can monitor the remaining clients via WMI.
Is anyone using a php front end with nagios, I saw a couple on the web, and read that hopefully Nagios 4.0 will be entirely php and support with sql. This would make configurations much easier to manage, although I have no problems working through putty, and correcting my typos in the code every few minutes.
I am impressed with the basic setup and would like to get into the distributed mode. It would be nice to have an agent that can be scheduled to call back once in a while to the Nagios server, in remote network environments.
I saw a nice little addon that monitors window clients via WMI. You only have to install the agent on an IIS webserver, and you can monitor the remaining clients via WMI.
Is anyone using a php front end with nagios, I saw a couple on the web, and read that hopefully Nagios 4.0 will be entirely php and support with sql. This would make configurations much easier to manage, although I have no problems working through putty, and correcting my typos in the code every few minutes.
I am impressed with the basic setup and would like to get into the distributed mode. It would be nice to have an agent that can be scheduled to call back once in a while to the Nagios server, in remote network environments.
-
- New Member
- Posts: 1
- Joined: Fri Oct 17, 2008 1:35 pm
- Location: San Francisco, CA, USA
Active check performance and limitations
Greetings,
I am curious as to what kind of performance numbers you are seeing on active checks for each DNX node, and if you have seen improvements with Nagios 3.0, which you should. In Nagios 3.0, there are two improvements that should lead to faster active check handling. (1) you can configure checks to be executed from a child process instead of a grandchild process, thereby saving you a fork(), and (2) the check results queue is no longer in memory, but it resides on disk, which means that, since there is a lighter memory image, the time to execute each fork() is commensurately reduced.
Best,
Emmanuel Feinsmith
Senior Software Architect and Engineer
Groundwork Open Source
I am curious as to what kind of performance numbers you are seeing on active checks for each DNX node, and if you have seen improvements with Nagios 3.0, which you should. In Nagios 3.0, there are two improvements that should lead to faster active check handling. (1) you can configure checks to be executed from a child process instead of a grandchild process, thereby saving you a fork(), and (2) the check results queue is no longer in memory, but it resides on disk, which means that, since there is a lighter memory image, the time to execute each fork() is commensurately reduced.
Best,
Emmanuel Feinsmith
Senior Software Architect and Engineer
Groundwork Open Source
-
- Member
- Posts: 278
- Joined: Wed Mar 21, 2007 9:12 am
Groundwork
I have had good luck with Groundwork. They have a community version that is free. It looks to be a pain to install, so I went with the VMWare image.novasam wrote: Is anyone using a php front end with nagios, I saw a couple on the web, and read that hopefully Nagios 4.0 will be entirely php and support with sql. This would make configurations much easier to manage, although I have no problems working through putty, and correcting my typos in the code every few minutes.
I am impressed with the basic setup and would like to get into the distributed mode. It would be nice to have an agent that can be scheduled to call back once in a while to the Nagios server, in remote network environments.
Groundwork has powerful templates and a nice UI that helps take some of the mystery out of Nagios. It also includes things like NagVis and Weathermap.
My environment is small enough that I don't need the distributed stuff, so I can't vouch for how well that works with the Groundwork framework.
http://www.groundworkopensource.com/