August 2014 - Xymon - Xymon mailman web

Setting up custom graph to monitor mail queue on postfix
by hobbit＠weiser.dk 12 Aug '14

12 Aug '14

Hi, I using Xymon to monitor a postfix mail server, and want to monitor the different mail queues on it. I have found this script: https://wiki.xymonton.org/doku.php/monitors:postfix I have the data comming from the mail server to Xymon server: Mails active: 3 amount equal to last measure. Mails bouncing: 0 amount equal to last measure. Mails in deferred State: 63 amount equal to last measure. Corrupt Mails: 0 amount equal to last measure. Incoming Mails: 0 amount equal to last measure. But I have not succeded in getting the graph to work, I have follow the steps on the wiki site above, and I have also gone though the help file, about costum graphs, in Xymon web site. In xymonserver.cfg have I added: - 'postfix=ncv' to the TEST2RRD string - NCV_postfix="*:GAUGE" Also in graph.cfg have I added: [postfix] FNPATTERN postfix,(.*).rrd TITLE Postfix statistics YAXIS Mail(s) DEF:p at RRDIDX@=@RRDFN@:lambda:AVERAGE LINE2:p at RRDIDX@#@COLOR@:@RRDPARAM@ GPRINT:p at RRDIDX@:LAST: \: %8.1lf (cur) GPRINT:p at RRDIDX@:MAX: \: %8.1lf (max) GPRINT:p at RRDIDX@:MIN: \: %8.1lf (min) GPRINT:p at RRDIDX@:AVERAGE: \: %8.1lf (avg)\n But it do not generate the rrd files, I think I have missed something, but I can't figure it out, I hope You can... Regards, Kim

6 9

cleint not reporting to hobbit server
by usaims＠yahoo.com 12 Aug '14

12 Aug '14

Greetings, I have this one linux client that is not reporting 'disk', 'cpu', 'ssh' and the custom external script raid-monitor to the hobbit server. What is reporting is 'conn', 'info', and 'trends'. I installed the client with 'apt-get install hobbit-client' on a ubuntu server. On the hobbitclient.log, it indicates: 2014-08-12 09:16:27 Cannot determine IP address of message recipient 10.218.275.25 2014-08-12 09:16:27 Whoops ! bb failed to send message - Cannot resolve hostname /usr/lib/hobbit/client/bin/hobbitclient-linux.sh: 43: who: Permission denied /usr/lib/hobbit/client/bin/hobbitclient-linux.sh: 55: /sbin/ifconfig: Permission denied /usr/lib/hobbit/client/bin/hobbitclient-linux.sh: 64: /sbin/ifconfig: Permission denied 2014-08-12 09:21:27 Cannot determine IP address of message recipient 10.218.275.25 2014-08-12 09:21:27 Whoops ! bb failed to send message - Cannot resolve hostname On the clientlaunch.log, it indicates, 2014-08-12 08:51:18 hobbitlaunch starting 2014-08-12 08:51:18 Loading tasklist configuration from /etc/hobbit/clientlaunch.cfg 2014-08-12 08:51:18 Task raid-monitor terminated, status 2 On the hobbit server in the /etc/hobbit/bb-hosts, it indicates: xxx..xxx.xxx.xxx server01 #raid Will somebody lead me in the right direction, thanks in advance. usaims

4 5

Help: how to make testip default behavior?
by feld＠feld.me 11 Aug '14

11 Aug '14

Hello, We recently moved our DNS hosting infrastructure to DNS Made Easy. The price is great, the DNS resolution speed is amazing, and it has all the features one could hope for. Except one feature that happens to be a big bug in my opinion: Any query that would normally return an NXDOMAIN instead returns 54.235.159.97 and they direct you to a webpage to tell you that the domain doesn't exist. This was not discovered in our testing. I'm extremely unsettled by this. Turns out we have thousands of entries in hosts.cfg that don't resolve to real hostnames. This is mostly intentional; it's RFC1918 stuff on management VLANs. Unfortunately due to the lack of a "testip" entry Xymon is following this wildcard that is being injected by DNS Made Easy. I've started adding it in batches but this is very tedious. Can I tell xymon globally to behave as if "testip" was set everywhere? Thanks

3 3

new install issue
by kspringer＠innovateteam.com 09 Aug '14

09 Aug '14

I've got a stable server hosted on AWS and I'm trying to build a new server on my ESXi VM server hosted elsewhere. Both are using Xymon 4.3.17 and both are build on Ubuntu 14.04. I've got Xymon2 up and supposedly configured the same as Xymon1. It's displaying the basic colored webpage with alerts functioning. My problem is that when I click any colored icon to view the drilldown info I get this message. #!/bin/sh # This is a wrapper for the Xymon svcstatus.cgi script . /usr/lib/xymon/server/etc/cgioptions.cfg exec /usr/lib/xymon/server/bin/svcstatus.cgi $CGI_SVC_OPTS Signature - Kris I tend to think this is an apache2 issue but I can't figure out what it is. I'm comparing Xymon1 and Xymon2 config files and I'm not seeing anything wrong. Anybody have a clue for me? Thank you. ------------------------------------------------ Kris Springer

6 8

extend http timeout
by kspringer＠innovateteam.com 08 Aug '14

08 Aug '14

Where and how do I adjust the http timeout? I have some content tests that take a while to respond and I need to extend the time to around 30 seconds. Signature - Kris Thank you. ------------------------------------------------ Kris Springer

3 3

graph with secondary ordinate
by martin.flemming＠desy.de 08 Aug '14

08 Aug '14

Hi ! I know, it's more a rrd-question, but how could i implement a second ordinate on the right side of the graph e.g. the values of my Check are Waiting: 6369 Running: 4114 Available: 5788 Slot Usage: 71.0780926053905 % and i want to have on the right side the Slot-Usage with percent my graph definition looks like [sges] TITLE BIRD JOBS YAXIS Jobs DEF:running=sges.rrd:Running:AVERAGE LINE2:running#FFCCCC:Number of running jobs COMMENT:\n GPRINT:running:LAST:running Jobs \: %5.1lf%s (cur) GPRINT:running:MAX: \: %5.1lf%s (max) GPRINT:running:MIN: \: %5.1lf%s (min) GPRINT:running:AVERAGE: \: %5.1lf%s (avg)\n DEF:available=sges.rrd:Available:AVERAGE LINE2:available#000099:Number of available Slots COMMENT:\n GPRINT:available:LAST:available Slots \: %5.1lf%s (cur) GPRINT:available:MAX: \: %5.1lf%s (max) GPRINT:available:MIN: \: %5.1lf%s (min) GPRINT:available:AVERAGE: \: %5.1lf%s (avg)\n DEF:usage=sges.rrd:SlotUsage:AVERAGE LINE2:usage#8C1717:Usage in % COMMENT:\n GPRINT:usage:LAST:Usage \: %5.1lf%s (cur) GPRINT:usage:MAX: \: %5.1lf%s (max) GPRINT:usage:MIN: \: %5.1lf%s (min) GPRINT:usage:AVERAGE: \: %5.1lf%s (avg)\n Thanks & cheers martin

1 1

sending results to another server
by Ejones＠egov.com 07 Aug '14

07 Aug '14

I just built a new xymon server on a new vlan - it will eventually replace my existing xymons on the old vlan. There are two xymons in different data centers. I'm sending the results of each to the other's page with the summary command - summary A-xymon.All 1.2.3.4 http://A-xymon/xymon/ works great works great. But trying to send the summary to the new xymon on the new vlan, not getting through. summary A-xymon.All 5.6.7.8 http://A-xymon/xymon/ fails I'm assuming this is a firewall issue - but I don't know the route this is taking. Would it be coming in on port 1984 like the regular xymon traffic?

2 1

Acknowledge alert doesn't work on pages with page name longer than 18 characters.
by lenko99＠gmail.com 07 Aug '14

07 Aug '14

Hi, I have run into issues with acknowledging alerts on pages where page name is longer than 18 characters. The ack page is just empty - it doesn't show any form. The acknowledge.cgi script throws "malloc(): memory corruption" error when its error output is redirected to some file. I have track it down to memory allocation in acknowledge.c on line 292: "re = (char *)malloc(8 + strlen(pagename));" I believe that allocating "(8 + strlen(pagename))" of memory for "re" is not enough if we put the pagename into it twice in the next line "sprintf(re, "%s$|^%s/.+", pagename, pagename);". I'm not quite sure why it works for names shorter than 19 chars though. Following change to acknowledge.c seems to fix the issue: --- acknowledge.c 2014-01-27 16:56:50.000000000 +0000 +++ acknowledge.c.new 2014-08-07 13:01:03.000000000 +0100 @@ -289,7 +289,7 @@ pcre *dummy; char *re; - re = (char *)malloc(8 + strlen(pagename)); + re = (char *)malloc(8 + strlen(pagename)*2); sprintf(re, "%s$|^%s/.+", pagename, pagename); dummy = compileregex(re); if (dummy) { This was tested on 64 bit Ubuntu 14.04 running xymon 4.3.17. Regards, Martin

1 0

External Powershell script not working
by schultmc＠ivytech.edu 07 Aug '14

07 Aug '14

We've created an external powershell script for monitoring one of our Exchange servers. The script has been listed in the externals section of BBwin.cfg and we've verified the script is outputing to C:\Program Files (x86)\BBWin\tmp but the results aren't showing up on the xymon server page. We have other servers running external scripts but haven't been able to determine the difference between the working external script and the non-working one. Powershell script: Add-PSSnapin Microsoft.Exchange.Management.PowerShell.Admin -ErrorAction SilentlyContinue Add-PSSnapin Microsoft.Exchange.Management.PowerShell.Support -ErrorAction SilentlyContinue $count = Get-TransportServer | Get-Message | Measure-Object | Select-Object Count -ExpandProperty Count $sAlarm = "green" $datetime = (Get-Date).ToString() If ($count -gt 150) { $sAlarm = "yellow" } $sAlarm,$datetime | Out-File -FilePath "C:\Program Files (x86)\BBWin\tmp\exchqueues" Output: green 8/6/2014 10:32:22 AM BBwin.cfg snippet: <externals> <setting name="timer" value="5m" /> <setting name="logstimer" value="60" /> <load value="C:\Windows\system32\WindowsPowerShell\v1.0\powershell.exe -nologo -ExecutionPolicy Unrestricted -File 'C:\Program Files(x86)\BBWin\ext\exchqueues.ps1'" />  </externals> -- ------------------------------------------------------------- Michael Schultheiss, RHCSA, RHCE OIT Administration Sr. Unix Systems Engineer Ivy Tech Community College schultmc at ivytech.edu

3 3

xymon-rclient
by jlaidman＠rebel-it.com.au 06 Aug '14

06 Aug '14

Firstly, please cc the mailing list. Secondly, best to upgrade to Xymon if that's possible. On 06/08/2014 9:54 PM, "Rod Simioni" <rod.simioni at gmail.com> wrote: > I'll need more flexiability because this xml requires custom headers, etc. I've created a custom script -- I placed it in the /server/ext dir and populated my :hobbitlaunch.cfg with > [foo] > ENVFILE $BBTHOME/etc/hobbitserver.cfg > CMD $BB/ext/myscript.sh > LOGFILE $BB/logs/myscript.log > INTERVAL 5m I can't tell what your script does so I'm guessing that it has the target server name hard coded. > I've placed 'foo' in the bb-hosts file for the bb server This is not needed. Generally, test names in bb-hosts are used only by the "network" tests (xymonnet or bbnet). Everything else uses their own configurations, typically in-script variables or separate configuration files. It's possible for a script to fetch it's configuration elements from bb-hosts, but that's not very common, and requires the script to do some extra work to get it. Again this is something I can't tell because I don't know what your script is doing. > but I want to monitor the client-less vips. If I placed 'foo' next to my client-less vips, it does nothing, do you have a clue? The only place you need to define your test name is on the script that sends the status message. Everything else is incidental or superfluous. Even the [foo] in hobbitlaunch.cfg is arbitrary, and as long as its unique, it can be any alphanumeric string of reasonable length. There's no reason to list the test name in the bb-hosts file. Your script needs to know the host name(s) it's going to test. Can you show the script? Have you looked at myscript.log? J

4 5