January 2011 - Xymon - Xymon mailman web

xymongen and xymonnet erratic
by losburn＠thepinnaclegroup.com 07 Jan '11

07 Jan '11

Env: 4.3.0 beta3 on RHEL5 Over a 24 hour period, the xymonnet process spends about 1/3 of its time in purple status, and 2/3 in green. This proportion holds true over a 7 day period too. Xymongen is very similar, with ratios of about 1/4th in purple, 3/4ths in green. When xymongen goes purple, the main web page stops updating, causing people to lose faith in the veracity of the data being reported, leading them to ignore xymon entirely. What log file(s) will be helpful? Thanks!

2 1

Display multiple values(comming from data base via SQL query) in a single page.
by jayantak007＠gmail.com 06 Jan '11

06 Jan '11

Hi All, I would like to display multiple values(comming from data base via SQL query) in a single page. I.e, how can i configure my xymon page as follows: *Xymon* *Current non-green Systems* *Thu Jan 06 08:41:38 2011* ------------------------------ *All Monitored Systems OK* 3 events received in the past 132 minutes Thu Jan 6 07:34:15 2011 home.scheel.dk http [image: red]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=home.scheel.dk&SERVICE=ht…> [image: >From -> To] [image: green]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=home.scheel.dk&SERVICE=ht…> Thu Jan 6 06:34:09 2011 tyge.sslug.dk http [image: clear]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=tyge.sslug.dk&SERVICE=htt…> [image: >From -> To] [image: green]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=tyge.sslug.dk&SERVICE=htt…> Thu Jan 6 06:29:08 2011 tyge.sslug.dk http [image: green]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=tyge.sslug.dk&SERVICE=htt…> [image: >From -> To] [image: clear]<http://www.xymon.com/xymon-cgi/historylog.sh?HOST=tyge.sslug.dk&SERVICE=htt…> No events acknowledged in the last 240 minutes With Regards, Jayanta.

1 0

Repeat: Issues with DOWNTIME, purple status, and sslcert test
by novosirj＠umdnj.edu 05 Jan '11

05 Jan '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I asked this question in about March of last year and didn't get any replies. Figured I'd give it another shot as it has been a persistent annoyance since then (though the behavior has changed slightly, which I've edited below): - --- Hunted around on Google and the mailing list archives about this one, but I'm turning nothing up. I have some downtime statements listed for a few hosts (names and IP's changed to protect probably nothing, but hey): 10.11.12.13 host1.umdnj.edu # "DOWNTIME=ldap,ldaps,sslcert:*:2200:2300:Development/Test Systems Cold Bacukup" 10.11.12.14 host2.umdnj.edu # "DOWNTIME=ldap,ldaps,sslcert:*:2200:2300:Development/Test Systems Cold Bacukup" 0.0.0.0 host3.umdnj.edu # "DOWNTIME=http,sslcert:*:2200:2300:Development/Test Systems Cold Bacukup" 0.0.0.0 host4.umdnj.edu # "DOWNTIME=http,sslcert:*:2200:2300:Development/Test Systems Cold Bacukup" The downtime is picked up fine, and correctly, in the "info" test on the hosts. What's interesting is what happens when the time rolls around with the sslcert test (I'll pick Friday night out of the air, but there's nuance every day): host1: Green until purple at 2245. host2: Ditto. host3: Ditto. host4: Blue at ~2200, red at 2245. All four share the same DOWNTIME definition, aside from the tests involved, so I'm mighty confused. Is there some interplay here that I'm not aware of (either by what triggers an sslcert test, https/ldaps/etc., or by using 0.0.0.0 on some of the hosts, the actual status of the hosts during the DOWNTIME, etc.) that someone might have some hints for teasing out? - -- - ---- _ _ _ _ ___ _ _ _ |Y#| | | |\/| | \ |\ | | |Ryan Novosielski - Sr. Systems Programmer |$&| |__| | | |__/ | \| _| |novosirj at umdnj.edu - 973/972.0922 (2-0922) \__/ Univ. of Med. and Dent.|IST/CST-Academic Svcs. - ADMC 450, Newark -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.14 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0kkYMACgkQmb+gadEcsb7WnwCcDjjcoVeCc40pBoQYhqBUo/oO pHkAnj2idab920W2cvFfsuT6Q7z/5bEV =Xk75 -----END PGP SIGNATURE-----

3 4

client-local.cfg
by doctor＠makelofine.org 04 Jan '11

04 Jan '11

Hi I'm running xymon client and server 4.3.0 beta3 on a server called toto, a linux host. In ~xymon/server/etc/client-local.cfg I created a section as follow : [toto] file:`find /test -name tata` On client side, in ~xymon/client/tmp/logfetch.toto.cfg, there is nothing like I configured in the client-local.cfg. It seems the server transfered the [linux] section, not the [toto] one. I tried to remove the logfetch.toto.cfg file, restarted both client and server with no effect. How could I correct this ? Regards and best wishes !

3 4

Graph stopped working
by joe.acquisto＠gmail.com 04 Jan '11

04 Jan '11

Using Hobbit 4.20, had a custom graph working fine. Copied that graph definition, and renamed the [subject] and the data names under it. When I saved it, not having changed a thing on the original definition, it disappeared from the page, replaced with the infamous ncv:name-of-graph. I commented out the new definition and saved hobbitgraph.cfg. No change. Deleted new def and saved. No change. So . . . umm, ??

2 3

Request for get data from Oracle database.
by jayantak007＠gmail.com 04 Jan '11

04 Jan '11

Hi All, I would like get data from oracle database on a particular table. For example: SELECT count(1) from emp; I want to display the result of the above query into my xymon page. Kindly let me know what the steps are for do the same. Which files need to modify/configure etc. With Regards, Jay.

3 3

hobbit-alerts.cfg and where's the needle
by sholmes42＠mac.com 04 Jan '11

04 Jan '11

Is there any way to find out exactly where a recipient is being picked up to be alerted for a specific host? I.e. (on Xymon 4.2.3) I have 2 recipients for all tests on a pair of hosts and there should be only one. They are windows application servers and both the app admins and unix admins are getting the alerts. The info pages for both hosts on that page show both groups as recipients. I've scoured the hobbit-alerts.cfg file and I can NOT see any way the unix admin group should be included as a recipient for either of the hosts on the page in question, nor for either of the individual hosts. What I'm looking for is a tool that will read the hobbit-alerts.cfg file and take args like a host name and recipient, and return an explanation of how the results are compiled, in particular, which config line or lines would result in a specific recipient being included. I know this is highly unlikely to exist, but I thought I'd ask anyway. Thanks! Steve Purdue University/ITSO

4 4