Xymon 4.3.5 released
list Henrik Størner
Hi, I have uploaded Xymon 4.3.5 to Sourceforge a few minutes ago, and it will start mirroring out soon. Source-file is available, and Debian packages for amd64 and i386 architectures. New stuff in this release: * the new "delayred"/ "delayyellow" options in hosts.cfg which will replease the "badTEST" settings that only worked for network tests; * the enhancements to the Critical Systems view so you can also see the eventlog for the hosts Performance-wise this release changes a couple of CGI-scripts so they don't have to load the entire hosts.cfg file, which can be quite heavy if you have a large hosts.cfg which may even be split out over lots of included files. And there are some nasty bugs squashed that could trigger crashes in both xymond and some of the webpage tools. The changes in this release: * Fix crash in CGI generating the "info" status column. * Fix broken handling of IGNORE for log-file analysis. * Fix broken clean-up of obsolete cookies (no user impact). * Devmon RRD handler: Fix missing initialisation, which might cause crashes of the RRD handler. * Fix crashes in xymond caused by faulty new library for storing cookies and host-information. * Fix memory corruption/crash in xymond caused by logging of multi-source statuses. * New "delayred" and "delayyellow" definitions for a host can be used to delay change to a yellow/red status for any status column (replaces the network-specific "badFOO" definitions). * analysis.cfg and alerts.cfg: New DISPLAYGROUP setting to select hosts by the group/group-only/group-except text. * New HOSTDOCURL setting in xymonserver.cfg. Replaces the xymongen "--docurl" and "--doccgi" options, and is used by all tools. * xymond_history option to control location of PID file. * Critical Systems view: Optionally show eventlog for the hosts present on the CS view. * Critical Systems view: Multiple --config options can now be used, to display critical systems from multiple configurations on one page. * Detailed status display: Speedup by no longer having to load the hosts.cfg file. * xymongen and xymonnet: Optionally load the hosts.cfg from xymond instead of having to read the file. The bug relating to DNS tests not timing out - see this thread: http://lists.xymon.com/archive/2011-August/032306.html - has not been fixed in this release. I need to do some more digging into why the C-ARES library doesn't timeout the way it should, and what can be done to work around / fix it. Regards, Henrik
list Tom Shimada
Hello, I am currently trying to create graphs separately from xymon to provide stats for our customer. We do not want to let them view our entire network. I realize that when i did a rrdtool dump for memory or disk, there are separate rrd databases but the DS name is the same. Ofcourse, you cannot have the same DS name when you rrdtool graph etc.... Could you tell me how I can create the same graphs on xymon? Thank you, Tom
list Vernon Everett
Hi Tom
We encountered a similar situation some time back, where we only wanted to
provide specific views to specific groups.
The easiest way of doing that, is to create another Xymon page per "view".
On that page, you have duplicates hosts entries of the servers you want them
to see, and use group-only to list the tests that will be visible to the
group.
You then use the security features of Apache to limit the group's access to
that page only.
Hope that helps.
Regards
Vernon
▸
On 12 October 2011 02:31, Tom Shimada <user-a4d567369965@xymon.invalid> wrote:
Hello, I am currently trying to create graphs separately from xymon to provide stats for our customer. We do not want to let them view our entire network. I realize that when i did a rrdtool dump for memory or disk, there are separate rrd databases but the DS name is the same. Ofcourse, you cannot have the same DS name when you rrdtool graph etc.... Could you tell me how I can create the same graphs on xymon? Thank you, Tom
list Tom Shimada
Hi Vernon, Thanks for the quick reply!! Yes, this does help but later today I was able to figure out how to output the graphs strictly using rrdtool. I do like the idea though! But I also need to add a custom graph for apache. The client wants an apache log report and i'm trying to pull that info and display it. Also, would you know how to set an alert when there is no http request for over 10 minutes? I'm having trouble finding out how to do this. I see some plug-ins for nagios, but I'm not familiar with it. Thanks in advance! Tom
▸
On Oct 11, 2011, at 3:18 PM, Vernon Everett wrote:
Hi Tom
We encountered a similar situation some time back, where we only wanted to provide specific views to specific groups.
The easiest way of doing that, is to create another Xymon page per "view".
On that page, you have duplicates hosts entries of the servers you want them to see, and use group-only to list the tests that will be visible to the group.
You then use the security features of Apache to limit the group's access to that page only.
Hope that helps.
Regards
Vernon
On 12 October 2011 02:31, Tom Shimada <user-a4d567369965@xymon.invalid> wrote:
Hello,
I am currently trying to create graphs separately from xymon to provide
stats for our customer. We do not want to let them view our entire network.
I realize that when i did a rrdtool dump for memory or disk, there are separate rrd databases
but the DS name is the same.
Ofcourse, you cannot have the same DS name when you rrdtool graph etc....
Could you tell me how I can create the same graphs on xymon?
Thank you,
Tom
list Henrik Størner
▸
On Wed, 12 Oct 2011 00:57:24 -0700, Tom Shimada <user-a4d567369965@xymon.invalid> wrote:
Also, would you know how to set an alert when there is no http request for over 10 minutes? I'm having trouble finding out how to do this.
The simplest way would be to have the Xymon client on the webserver monitor the Apache "access" log file. Then you can configure analysis.cfg on the Xymon server to trigger an alert if that file is more than 600 seconds old, with FILE /var/log/apache/access.log mtime<600 Regards, Henrik
list Tom Shimada
Got it!!! Thank you so much for the help! Tom
▸
On Oct 13, 2011, at 5:36 AM, user-ce4a2c883f75@xymon.invalid wrote:
On Wed, 12 Oct 2011 00:57:24 -0700, Tom Shimada <user-a4d567369965@xymon.invalid> wrote:Also, would you know how to set an alert when there is no http request for over 10 minutes? I'm having trouble finding out how to do this.The simplest way would be to have the Xymon client on the webserver monitor the Apache "access" log file. Then you can configure analysis.cfg on the Xymon server to trigger an alert if that file is more than 600 seconds old, with FILE /var/log/apache/access.log mtime<600 Regards, Henrik