January 2013 - Xymon - Xymon mailman web

Xymon metric data into Graphite?
by cleaver＠terabithia.org 07 Jan '13

07 Jan '13

> I see that there is a xymond_rrd "--processor=/path/to/script" option. > In theory, could that be used to send all metric data to graphite? > If so, is there also an option to disable Xymon RRD graphs? > If I had to, I could probably create some sort or RAM disk for RRD graphs > and have them contain 1 metric each, but would be nice to just disable RRD. > I found people have created monitoring of Graphite files and send results > to Xymon, but I want to store Xymon metric data in Graphite. > > Thanks > Craig That's exactly what you'd want to do; a while back I was running an RRD processor pipe into an OpenTSDB instance running locally (the processor fixup script did some trivial, and cheap, rewriting) the same way. In terms of preventing xymond_rrd from doing the actual writing of rrd files (even though you're using it for raw data parsing still), I agree that that's annoying. I've only done cursory testing on it, but the attached patch adds a ---no-rrd option to xymond_rrd, which should do what you want. Regards, -jc

1 0

Bug / Feature request rrd files heartbeat.
by Michael.Beatty＠sherwin.com 07 Jan '13

07 Jan '13

Due to network limitations and the number of hosts involved, I decided to back the frequency of reports for the clients back to 15 mintues. I updated the INTERVAL parm in clientlauch.cfg.... easy enough. However, now, none of my graphs are working. Reason being, all of my rrd files have heartbeats of 10 mintues. After looking through the code, I was disappointed to see that the 600 second heartbeat is hardcoded into all of the /xymond/rrd/do_*.c files. This leaves me with two options. 1. Recode all of those programs to create the RRD's from a variable. Or 2. Run rrdtool tune on all of rrds after they are created. Or am I missing something? I would like to see that the heartbeat for rrds that are created would always be equal to 2 X Client INTERVAL. Also, it would be helpful to have in the documentation that you can set the rrd STEP parameter in rrddefinitions.cfg -- Michael Beatty

1 0

connection refused when running "xymon drop"
by ciprian.parfon＠gmail.com 07 Jan '13

07 Jan '13

Hi I've tried to drop "conn" check for two hosts and I'm getting "connection refused" (saw it in xymond.log). I've tried from both the server and client. 2013-01-07 12:35:13 Refused message from xx.xx.xx.xx: drop dev.domain.com conn 2013-01-07 12:37:20 Refused message from xx.xx.xx.xx: drop dev.domain.com conn Any thoughts ? --

3 3

is noconn ignored ?
by ciprian.parfon＠gmail.com 07 Jan '13

07 Jan '13

Hi Gents, Need a bit of help on something. In my hosts.cfg on fresh installed xymon server (v 4.3.10) I have this: group <H4>Websites</H4> group-compress <H5>Main</H5> 127.0.0.1 domain_1.com # noconn http://www.domain_1.com 127.0.0.2 domain_2.com # noconn https://www.domain_2.com group-compress <H5>Redmine</H5> 127.0.0.3 dev.managinglife.com # noconn https://dev.domain_1.com/redmine/ group-compress <H5>Jenkins</H5> 127.0.0.4 dev.domain_1.com # noconn http://dev.domain_1.com/jenkins/ 127.0.0.4 preprod.domain_2.com # noconn http://preprod.domain_2.com/jenkins/ 127.0.0.6 prod.domain_2.com # noconn http://prod.domain_2.com/jenkins/ Although for the first 2 websites all looks good, for the ones which follow below, I get the "conn" column which shows "red" (that's because those hosts don't respond to ping, being default ec2 instances). Any idea how can I remove the "conn" column/check ? Thanks! -- Ciprian Parfon System & Network Engineer +40 721879113 ciprian.parfon at gmail.com

4 3

Xymon metric data into Graphite?
by craig＠cookitservices.com 07 Jan '13

07 Jan '13

An HTML attachment was scrubbed... URL: <http://lists.xymon.com/pipermail/xymon/attachments/20130105/581c5d0c/attach…>

2 1

Again netapp.pl question
by linuxmasterjedi＠free.fr 06 Jan '13

06 Jan '13

Hi, I'm (still) using netapp.pl 1.21 on XYmon 4.2.2 et Data ONTAP 8.1.1P2 When my netapp.pl runs correctly, my monitored netapp has around 20 columns of green/yellow/red status icons. Sometimes, all columns are dropped except "snaplist, snapmirror, sysconfig, uptime, volume". I've planned the test each 10 min. Do you have any ideas why most of the column are 'dropped' and then, reappear later ? Thanks.

1 0

Netapp.pl & Cluster
by linuxmasterjedi＠free.fr 06 Jan '13

06 Jan '13

Hi, I'm running netapp.pl

1 1

Netappl.pl and Cluster
by linuxmasterjedi＠free.fr 06 Jan '13

06 Jan '13

Hi, I'm using netapp.pl 1.21 on XYmon 4.2.2. We recently change the wiring at the back of our netapp, then upgraded and rebooted them. I'm running Data ONTAP 8.1.1P2 Now, the plugin netapp.pl display yellow (warning) with this message : Controller Failover enabled, MYNetApp01 is up. current time: 06 Jan2013 07:52:14 UP 15:20:36, partner 'MYNetApp02', cluster monitor enabled VIA Interconnect is up (link 0 up, link 1 up), takeover capability on-line partner update TAKEOVER_ENABLED (06 Jan2013 07:52:13) Any ideas why is it Yellow instead of Green ? Thanks

1 0

How to submit patches for Xymon?
by tschmidt＠micron.com 04 Jan '13

04 Jan '13

I have been working with Xymon 4.3.10 for a while now as an upgrade from Big Brother. In doing this, I have found a couple bugs and added a few enhancements, and want to submit these patches to get them rolled into the next Xymon update. How do I get these submitted? Do I just add the patch as an E-mail attachment here? Thanks...Tom -- Tom L. Schmidt Corporate Engineering Principal System Administration Lead Micron Technology, Inc. http://www.micron.com/ 8000 S. Federal Way P.O. Box 6 Mail Stop 01-351 Boise, Idaho USA 83707-0006 mailto:tschmidt at micron.com http://peunix.micron.com/

2 1

Emails not being sent
by Martin.Ward＠colt.net 04 Jan '13

04 Jan '13

Hi all, My Xymon server is set up so that I have a number of pages for the support teams to monitor (the usual Xymon pages) but I also have a subset of pages that have a cut-down list of servers and monitor items that I want the 24x7 team to look at and be alerted on. I did this by simply duplicating the main pages, changing the IP address to 0.0.0.0 and removing any items I didn't want alerted on (group-execpt is my friend here!). I also configured the alerts.cfg so that the support teams would get emails if things go wrong on their pages and the 24x7 team get alerts if things go wrong on theirs, obviously... one entry looks like this: Alerts.cfg: PAGE=SYS IGNORE HOST=%sys[12].lon.oss.colt.net SERVICE=procs,ports # These are reported by the sys-jb process SCRIPT /opt/xymon/server/ext/html_mail.pl support at colt.flibble.net DURATION>10m repeat=1d COLOR=yellow,red PAGE=IN-SYS IGNORE HOST=%sys[12].lon.oss.colt.net SERVICE=procs,ports # These are reported by the sys-jb process SCRIPT /opt/xymon/server/ext/html_mail.pl 24x7 at colt.flibble.net DURATION>1h repeat=8h COLOR=red RECOVERED The theory here is that if the disk column goes yellow then only support at colt.flibble.net gets and email whereas is the disk column goes red then both support at colt.flibble.net and 24x7 at colt.flibble.net get an email. Sadly only support at colt.flibble.net gets an email in both cases. Support does always get an email so I know that the actual tests work, I just cannot get the system to send out the second email! I've gone through every log file and large amounts of debug output to no avail. The output of xymond_alert -dump-config looks fine: 173 PAGE=SYS IGNORE REPEAT=30 HOST=%sys[12].lon.oss.colt.net SERVICE=procs,ports STOP SCRIPT /opt/xymon/server/ext/html_mail.pl support at colt.flibble.net FORMAT=SCRIPT REPEAT=1440 COLOR=yellow,red DURATION>10 176 PAGE=IN-SYS IGNORE REPEAT=30 HOST=%sys[12].lon.oss.colt.net SERVICE=procs,ports STOP SCRIPT /opt/xymon/server/ext/html_mail.pl 24x7 at colt.flibble.net FORMAT=SCRIPT REPEAT=480 COLOR=red DURATION>60 RECOVERED Suggestions are always appreciated, |\/|

2 2