June 2007 - Xymon - Xymon mailman web

[bug] bb2.html show red:nopropagate:
by charles.goyard＠orange-ftgroup.com 21 Jun '07

21 Jun '07

Hi, I encounter what looks much like a bug in bbgen. I have two hosts in failover, so I have the following logic : cwd-resa01o.cwd93.cvf.is_master = green cwd-resa02o.cwd93.cvf.is_master = red host-cluster = green (host1.is_master XOR host2.is_master with bbcombogen) So, either host1 or host2 is red all the time, and I don't want to propagate it. So I did : 10.193.173.21 cwd-resa01o.cwd93.cvf # ssh NOPROPRED:is_master,procs_master 10.193.173.22 cwd-resa02o.cwd93.cvf # ssh NOPROPRED:is_master,procs_master When cwd-resa02o.cwd93.cvf.is_master gets red, it does not show on bb2, but when cwd-resa01o.cwd93.cvf.is_master, it propagates and I get the following on bb2.html : is_master cwd-resa01o.cwd93.cvf is_master:red:nopropagate:2m It's somewhat strange to see "nopropagate" on the bb2 page... The debug log shows : load_bbhosts: -- got line '10.193.173.21 cwd-resa01o.cwd93.cvf # ssh NOPROPRED:is_master,procs_master' init_state(cwd-resa01o,cwd93,cvf.is_master, 135215936, ...) find_or_create_column(is_master) init_state: hostname=cwd-resa01o.cwd93.cvf, testname=is_master, color=5, acked=0, age=2m, oldage=0, propagate=0, alert=0 init_state(cwd-resa02o,cwd93,cvf.is_master, 135264320, ...) find_or_create_column(is_master) init_state: hostname=cwd-resa02o.cwd93.cvf, testname=is_master, color=5, acked=0, age=33s, oldage=0, propagate=0, alert=0 Host: cwd-resa02o.cwd93.cvf, ip: 10.193.173.22, name: cwd-resa02o.cwd93.cvf, color: 0, old: 0, anywaps: 0, wapcolor: -1, pretitle: '(NULL)', noprop-y: (NULL), noprop-r: ,is_master,procs_master,, noprop-p: (NULL), noprop-ack: (NULL), waps: (NULL) Test: is_master, alert 0, propagate 0, state 5, age: 33s, oldage: 0 Host: cwd-resa01o.cwd93.cvf, ip: 10.193.173.21, name: cwd-resa01o.cwd93.cvf, color: 4, old: 0, anywaps: 0, wapcolor: -1, pretitle: '(NULL)', noprop-y: (NULL), noprop-r: ,is_master,procs_master,, noprop-p: (NULL), noprop-ack: (NULL), waps: (NULL) Test: is_master, alert 0, propagate 0, state 5, age: 2m, oldage: 0 I run bbgen as: bbgen --recentgifs --subpagecolumns=2 --report --nopropred=procs_master,is_master I also tried without NOPROPRED and only --nopropred, same thing. May I say that I ran bbgen against a minimalistic bb-hosts file, with only these two hosts. Maybe that one already had been worked out ? -- Charles Goyard - charles.goyard at orange-ftgroup.com - (+33) 1 45 38 01 31 Orange Business Services - online multimedia // ingénierie

1 0

Local test limits
by Bart.Gillis＠msp.be 21 Jun '07

21 Jun '07

Hello, In the bb-hosts file you can set limits on the network test by specifying the badTEST x:y:z parameter. Is there a way that you can specify such limits for local tests (cpu, custom tests,...) to? If I set, for instance badcpu 1:2:3 as a parameter, this is not displayed in the configuration report and does not work: Network tests Service NK C/Y/R limits Specifics conn No 1/2/3 Local tests Service NK C/Y/R limits Configuration (NB: Thresholds on client may differ) cpu No -/-/- UNIX - Yellow: Load average > 1.5, Red: Load average > 3.0 Windows - Yellow: CPU utilisation > 80%, Red: CPU utilisation > 95% Best Regards Bart Gillis

2 1

Local test limits
by Bart.Gillis＠msp.be 21 Jun '07

21 Jun '07

Hello, In the bb-hosts file you can set limits on the network test by specifying the noTEST x:y:z parameter. Is there a way that you can specify such limits for local tests (cpu, custom tests,...) to? If I set, for instance nocpu 1:2:3 as a parameter, this is not displayed in the configuration report and does not work: Network tests Service NK C/Y/R limits Specifics conn No 1/2/3 Local tests Service NK C/Y/R limits Configuration (NB: Thresholds on client may differ) cpu No -/-/- UNIX - Yellow: Load average > 1.5, Red: Load average > 3.0 Windows - Yellow: CPU utilisation > 80%, Red: CPU utilisation > 95% Best Regards Bart Gillis

1 0

MAXMSG_*
by dbourque＠weatherdata.com 20 Jun '07

20 Jun '07

Hi, I set these values in the hobbitserver.cfg: MAXMSG_STATUS="1024" MAXMSG_CLIENT="1024" MAXMSG_DATA="1024" I still get some : Latest errormessages: Oversize data/client msg from n.n.n.n truncated (n=2003124, limit 1048576) If I got the client info dump on the web page for that client. It's always back to green when I go look at it. When it's green, if I cut and paste all the text into a text file and and run du -k on it, I get 80. How can that output possibly grow over twenty times to trigger an hobbitd warning ??? Am I not understanding exactly what those limits apply to ? Thanks! -Dan

1 1

Problem with multiple display servers and hobbit.
by sigurdur＠raforninn.is 20 Jun '07

20 Jun '07

Hiya. I am having a bit of a problem with multiple display servers and hobbit. I have a central server that is accessable from the internet and I have a couple of servers here and there to monitor devices for me. The devices can't send information to the central server, only the local server at each location. I configured each local server to have two bb displays in hobbitserver.cfg. (See below) # General settings BBPORT="1984" # Portnumber where hobbitd/bbd listens BBDISP="0.0.0.0" # IP of a single hobbit/bbd server BBDISPLAYS="$BBSERVERIP 11.11.11.11" # IP of multiple hobbit/bbd servers. If used, BBDISP must be 0.0.0.0 The second IP address is a fake one on purpose, however in my config it is real. And in the hobbitlaunch.cfg file the line that runs the daemon is the following. [hobbitd] ENVFILE /usr/lib/hobbit/server/etc/hobbitserver.cfg CMD hobbitd --pidfile=$BBSERVERLOGS/hobbitd.pid --restart= $BBTMP/hobbitd.chk --checkpoint-file=$BBTMP/hobbitd.chk --checkpoint-interval=50 --log=$BBSERVERLOGS/hobbitd.log --admin-senders=127.0.0.1 --store-clientlogs=!msgs This also is all good afaik. When I look at the local server, the conn status and everything is good all the time. However when I look at the central server the conn tab is good for 4 minutes or so and then every server that has a local server is down for 1. Has anyone experienced this problem and/or knows a solution to this? With regards, Sigurður Guðbrandsson.

2 2

proc monitor flag red an existing process
by sa＠hydre.org 20 Jun '07

20 Jun '07

Hi, I monitor the process tor with this line in hobbit-client.cfg : PROC "%^tor" TEXT=tor It works well except when the memory allocated for the process exceed 99999 : the monitor flag the process red... Perhaps it's due to a wrong regular expression or perhaps it's the shift in the PS output : 28894 28893 _ntp Mon05PM S 4 0.0 0:03.60 0.1 900 1344 ntpd: ntp engine (ntpd) 36112 1 tor Mon06PM S 20 0.0 233:44.10 9.6 98692 101272 tor 42051 1 fetchmail Fri06PM Ss 4 0.0 0:19.02 0.2 1760 3952 fetchmail -f /usr/local/etc/fetchmailrc -d 300 Any clue? Thanks Stéphane.

3 3

hobbitd_client: fatal: libpcre.so.0: open failed
by uma.bandi＠ge.com 20 Jun '07

20 Jun '07

Hi all, I am getting following error while I am running runclient.sh, could you please help me hobbitd_client: fatal: libpcre.so.0: open failed Thanks Uma Maheswar

4 4

Disabled systems view
by rsmrcina＠wi.rr.com 20 Jun '07

20 Jun '07

I'm not sure if this has come up before, but I have a customer where the network admins are finally getting the hang of disabling systems (instead of letting them sit red for days on end). It might be a good feature to have a display to see which systems are disabled. Maybe hang a menu item off the view menu, and display it like the non-green systems, but without the event log. -- Rich Smrcina VM Assist, Inc. Phone: 414-491-6001 Ans Service: 360-715-2467 rich.smrcina at vmassist.com Catch the WAVV! http://www.wavv.org WAVV 2008 - Chattanooga - April 18-22, 2008

5 8

problem with swap usage under Solaris
by Dominique.Frise＠unil.ch 19 Jun '07

19 Jun '07

Hi Henrik, I wrote about this issue several weeks ago and since I didn't get any answer, I assume it was not clear enough.(http://www.hobbitmon.com/hobbiton/2007/04/msg00119.html) Let'me give a second try ;-) We have several SAP servers under Solaris that are constantly over the swap warning threshold (50 %). SAP application does indeed *reserve* a lot of swap that confuses hobbit. In the following example the server has 32GB of RAM memory and 32GB of disk swap space. Swap outputs ------------ # swap -s total: 22347128k bytes allocated + 18080512k reserved = 40427640k used, 13640896k available # swap -l swapfile dev swaplo blocks free /dev/md/dsk/d5 85,5 16 66341232 65799424 Hobbit memory col. display -------------------------- Memory Used Total Percentage &green Physical 30460M 32768M 92% &yellow Swap 39493M 52779M 74% Top display ----------- last pid: 14928; load avg: 2.68, 2.37, 2.00; up 8+05:24:38 16:22:04 479 processes: 476 sleeping, 3 on cpu CPU states: 81.3% idle, 11.4% user, 7.3% kernel, 0.0% iowait, 0.0% swap Memory: 32G phys mem, 966M free mem, 32G total swap, 31G free swap ... ... Hobbit computes Used and Total using "swap -s" and this gives a percentage of 74% where "top" tool only show ~ 3% (1/32) Since version 3.6, "top" for Solaris uses output of swap -l for swap usage. (details under #24 of the faq at http://www.unixtop.org/faq.shtml#title6) We think Hobbit memory display should also be based on *disk* swap only (output of swap -l) Dominique UNIL - University of Lausanne

1 0

Option stop parsing hobbit-alerts.cfg when a rule is hit?
by chrisw＠supranet.net 19 Jun '07

19 Jun '07

Hello, I've been using hobbit for awhile, it's been working great. I was using BB, when I had originally ported my rules over, I thought that 'STOP' in hobbit-alerts.cfg meant "if this rule hits, stop checking all rules". Obviously it does not mean this. I see that this 'feature' has been discussed in the past: http://www.hswn.dk/hobbiton/2005/01/msg00298.html Has any progress been made on it, or is this simply a feature that's not requested frequently? We've got hundreds of hosts being monitored, we sometimes put rules in for specific hosts that we only want to alert certain people. Thanks for any feedback. --Chris

1 0