October 2005 - Xymon - Xymon mailman web

[hobbit] Ext script help needed.
by nfigaro＠effigie.fr 21 Oct '05

21 Oct '05

> -----Message d'origine----- > De : Thomas [mailto:tlp-hobbit at holme-pedersen.dk] > Envoyé : vendredi 21 octobre 2005 09:03 > À : hobbit > Objet : [hobbit] Ext script help needed. > > > Hi All, Hi, > > Please help me with this little problem I have, as it is > bothring be a lot. > > I have writte an ext script which logs into a FTP server does a dir > command and then excits. All FTP commands are specified in a > input file > and the output is saved in a temp file used for the status message to > Hobbit. > > Henrik kindly advised me to use a bbhostgrep to search the FTP server > details from the bb-hosts file and this works as expected. In > order to > speed up the testing I have created a "subscript" which I > call with som > paramaters from the "master script" and this way have some > parrallel run > of the test. Its like this: > > $BBHOME/ext/ftp-test.sh $HOST $USER $PASSWD & > > This also works fine. > > The problem starts when a FTP connection fails or is slow. Then my > subscript is not terminated and the master script hangs. This in turn > leads to purple alarms for all hosts. I have tried to implement in my > master script a check using ps for a previously running subscribt for > the individual hosts, but this is not work as expected. > > So my question is how can I have my "sub script" really fork from the > master script ? I do not need it to return to the master > script because > the subscript sends the status to Hobbit. Just a suggestion : You can write your sub script with perl and use Net::FTP module. This module includes a timeout option for ftp. If the connection is to slow or fails, the perl script can generate a specific message, and you'll send a red message to hobbit. I don't know if there is a timeout option to the standard ftp command. Nicolas Figaro > > Any help is greatly appreciated. > > Regards, Thomas > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk > > >

1 0

another approach: http check tweaking
by pbliss＠techvalleycom.com 21 Oct '05

21 Oct '05

Hello all, I'm having a bear of a time trying to figure out why my http checks are intermittently timing out on me. It will work fine for hours, and then I'll get an http timeout. It's been REALLY hard to catch it while it's happening because the "outage" lasts between 10 and 50 seconds. All this started after the Hobbit system was rudely rebooted so I'm reluctant to believe that there's a real problem. So my revised question, is the http monitor tweakable to be more permissive or am I barking up the wrong tree? Thanks in advance, Paul

2 1

Paging Problem
by Al.Jeffcoat＠orhs.org 20 Oct '05

20 Oct '05

Hello All, I've got paging setup and defined on my hobbit server. It has worked in the past, but at some point, stopped. Is there a way to test the sending of an alert? I know there's a command to test the alert rules, but is there a way to tell hobbit to send a page itself, as a test? I know that qpage is working, it works with the qpage -i function. The wrapper script is working (thanks Andy France) if I manually set the variables that it needs to execute($COLOR, $STATUS, etc), it sends a page. Hobbit thinks it is sending alerts based on the notications log file, but pages are not being sent. As a test, I have added a line to the wrapper script that just echoes to a log file when it is executed, but this file is still empty which tells me that it hasn't been executed by hobbit (it has been written to during the manual testing, I emptied the file for a fresh start). E-mails, however, are working fine. TIA, Al

1 0

[hobbit] Hobbit client
by nfigaro＠effigie.fr 20 Oct '05

20 Oct '05

> >So I also need information about installing up hobbit on a > AIX server (I have > >not tried to install hobbit yet on AIX, but I will try it next week). > > Hobbit should compile fine on AIX - I believe there are other > people using it on both AIX 4.3 and 5.<something>. I confirm hobbit works fine on AIX servers (aix 5.2). I'm still using a snapshot, but compiling the 4.1.2 release should be easy. rrdtool was harder to compile than hobbit on AIX. Nicolas Figaro > > > Regards, > Henrik > > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk

2 1

[hobbit] client disk graph differences
by Al.Jeffcoat＠orhs.org 20 Oct '05

20 Oct '05

I had this same problem, didn't notice it until adding a new client, it seemed some graphs were all mixed in where they had been split before. Some where split as usual. I am not sure if this was the fix, but I had stopped the hobbit server, cleaned $HOBBIT/tmp (all except the chk files), and then restarted. After this, all the graphs were back to being split out. HTH. Al -----Original Message----- From: Michael Nemeth [mailto:michael.nemeth at lmco.com] Sent: Tuesday, October 18, 2005 9:12 AM To: hobbit at hswn.dk Subject: Re: [hobbit] client disk graph differences Did you get an answer to this ? If you did can you tell me it! Ive got the same problem! Pat Vaughan wrote: >I don't know if any else had noticed this, but I just started >experimenting with the hobbit client in 4.1.2. I've found that the Hobbit >server generates the disk RRD graphs in seperate graphs of 4 filesystems >each. Which is a behavior I'd like to use in some of my custom graphs, >but I don't see how to set that. > >But, on the server with the new Hobbit client, all of the filesystems are >thrown into one big graph. With over 40 filesystems on this one server, >the graph isn't too usefull. > >Is there a setting somewhere to tell Hobbit to split up X rrd files into >seperate graphs? > > >To unsubscribe from the hobbit list, send an e-mail to >hobbit-unsubscribe at hswn.dk > > > > > -- +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ | _p_ Mike Nemeth | ___| |_____ email(w) michael.nemeth at lmco.com Work: 856 359-1425 |><___________) | Home Page:http://www.geocities.com/mjnemeth/ | Work Page:http://faraday.motown.lmco.com:3000/~nemethm/ | Work Page:http://ortsweb/~mnemeth/ |++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

4 6

[hobbit] http monitor is reporting servers down that actually aren't
by pbliss＠techvalleycom.com 20 Oct '05

20 Oct '05

Thank you very much Henrik, for your help. When I do: server:/# /usr/lib/hobbit/server/bin/bbcmd bbtest-net --debug HOSTNAME.FROM.BBHOSTS 2005-10-17 09:18:56 Using default environment file /usr/lib/hobbit/server/etc/hobbitserver.cfg server:/# That's all I get. I'm sorry if I'm just really new at Hobbit (which I am) but I'm not sure if that's the info it's supposed to return to me. In the actual error page when it's reporting an outage it says: * http://HOSTNAME.FROM.BBHOSTS/ - Server timeout Seconds: 10.33 The problem with trying to debug this issue is that the "outages" typically last about 9 seconds, so by the time I go to look at the Hobbit box, the "outage" is over. This only is effecting http monitored hosts and is not really specific to which one. Again, thank you so much to all who are taking the time to read this and help out. -Paul -----Original Message----- From: Henrik Stoerner [mailto:henrik at hswn.dk] Sent: Thursday, October 13, 2005 8:16 AM To: hobbit at hswn.dk Subject: Re: [hobbit] http monitor is reporting servers down that actually aren't On Thu, Oct 13, 2005 at 06:22:48AM -0400, Paul Bliss wrote: > Strangely, Hobbit is reporting that http is not functioning on several boxes that I know are functional and reachable. > I've looked through the logs and can't find any indication as to why they failed. What kind of error does Hobbit report (timeout, could not connect, bad HTTP code... ) ? > PS. This Hobbit server worked fine, until the server was accidentally cold-booted. > Any suggestions from the experts? Try running bbcmd bbtest-net --debug HOSTNAME.FROM.BBHOSTS This will do the network tests for this host and dump debugging output to the console. Henrik To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

1 0

[hobbit] Too many line on disk graph
by Al.Jeffcoat＠orhs.org 20 Oct '05

20 Oct '05

Hmmm... During the time that the graphs were "conglomorated" for me also, I too had disk events... Actually, I had a red and a yellow event. The yellow event was real, and the timing of fixing it coincided with cleaning the tmp directory, and restarting hobbit. Strange thing though, now that I am looking around. During this yellow disk event, hobbit also appeared to ignore the disk rules in hobbit-clients.cfg for another client, which caused a red event for this client. When the yellow event was fixed, the red was fixed also, and apparently the graphs. Al -----Original Message----- From: Pat Vaughan [mailto:patrick_a_vaughan at hotmail.com] Sent: Thursday, October 20, 2005 9:02 AM To: hobbit at hswn.dk Subject: Re: [hobbit] Too many line on disk graph Another possible clue, when I came in this morning there was a yellow state on the disk test and Hobbit was displaying one huge disk graph. I even did a reload on the page to make sure Hobbit had to regenerate it. After I fixed the error and the state went back to green, the graphs were back to normal. It might just be a coincidence, though. > A couple of us have had the same problem. Stopping the server, clearing > out the Hobbit server tmp subdirectory and restarting it was suggested. > > At first it didn't fix anything (and I did a forced reload on the disk web > page), but after a couple of minutes the graphs went back to being split > as normal. > > > >> OK the server seam to be no long spitting the disk graphs. This is >> true for all source of data, bb-clients local and remote >> clients,. This is with snapshot done on Oct. 16 night. >> >> -- >> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >> | _p_ Mike Nemeth >> | ___| |_____ email(w) michael.nemeth at lmco.com Work: 856 359-1425 >> |><___________) | Home >> Page:http://www.geocities.com/mjnemeth/ >> | Work Page:http://faraday.motown.lmco.com:3000/~nemethm/ >> | Work Page:http://ortsweb/~mnemeth/ >> |++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >> >> >> To unsubscribe from the hobbit list, send an e-mail to >> hobbit-unsubscribe at hswn.dk >> >> >> > > > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk > > > To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

1 0

Weird vmstat collection on FreeBSD client
by arnoud＠koekiemonster.com 20 Oct '05

20 Oct '05

I have Hobbit 4.1.2 server and client running on FreeBSD 5.3. I'm trying to get a nice vmstat graph out of it, but something is wrong. I just noticed the comments beneath the graph don't count up to 100%: System : 4.3 (cur) User : 90.3 (cur) Idle : 185.9 (cur) Looking at the client data these numbers don't come back any where: [vmstat] procs memory page disk faults cpu r b w avm fre flt re pi po fr sr ad0 in sy cs us sy id 1 12 0 526080 50484 367 0 0 0 358 9 0 347 1776 336 5 4 91 3 13 0 526064 49920 301 0 0 0 293 0 3 349 1591 325 2 4 95 Where can I check what data is collected and how it is parsed to rrd, or turn the debugging on? I found the grapher.cgi and rrd files are not readable. Arnoud Post

1 0

Too many line on disk graph
by michael.nemeth＠lmco.com 20 Oct '05

20 Oct '05

OK the server seam to be no long spitting the disk graphs. This is true for all source of data, bb-clients local and remote clients,. This is with snapshot done on Oct. 16 night. -- +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ | _p_ Mike Nemeth | ___| |_____ email(w) michael.nemeth at lmco.com Work: 856 359-1425 |><___________) | Home Page:http://www.geocities.com/mjnemeth/ | Work Page:http://faraday.motown.lmco.com:3000/~nemethm/ | Work Page:http://ortsweb/~mnemeth/ |++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

4 8

[hobbit] Hobbitclient problem and Confreport problem
by Mark.Deiss＠acs-inc.com 20 Oct '05

20 Oct '05

If you really need a longer line, trying using: ps -efww The deadcat offerings sometimes use: ps auxw or ps auxww I am curious if anyone has noticed any problems pumping full process lists from N servers to a hobbit server. We have boxes running a lot of java processes with really lonnnngggggg command lines. The single "w" in the ps command should help truncate some of the line length. If the process monitoring does not require processing the entire lonnngggg line, then single "w" would be better the "ww". Keep in mind the lonnngggg lines also make for a rather congested web display for the process list. -----Original Message----- From: Francesco Duranti [mailto:fduranti at q8.it] Sent: Wednesday, October 19, 2005 7:14 AM To: hobbit at hswn.dk Subject: [hobbit] Hobbitclient problem and Confreport problem Hi all... I'm starting to implement some checks on disks and process and i'm getting some problems... I'm running the yesterday snapshot on RHEL3 server and client. >From what i see i have a problem with the ps -efw on the linux client, it will display something more then ps -ef but i don't get the entire ps rows... For example: ps -ef qmaild 22356 4850 0 Aug10 ? 00:02:29 /usr/local/bin/tcpserver -v -R - ps -efw qmaild 22356 4850 0 Aug10 ? 00:02:29 /usr/local/bin/tcpserver -v -R -l cl-mail -x /prodQ8/qmail/tcp/tcp.smtp.cdb-c 100 - ps -ef --width=1000 qmaild 22356 4850 0 Aug10 ? 00:02:29 /usr/local/bin/tcpserver -v -R -l cl-mail -x /prodQ8/qmail/tcp/tcp.smtp.cdb-c 100 -u 49921 -g 41420 0 smtp /prodQ8/qmail/bin/qmail-smtpd Another thing with confreport scripts. In the hobbit-client.cfg file I have: HOST=cl-mail PROC "%^qmail-send$" 1 1 PROC "%/prodQ8/qmail/bin/qmail-smtpd" 1 1 PROC "%/prodQ8/qmail/bin/qmail-qmqpd" 1 1 DISK "/prodQ8/qmail" 5 10 HOST=itromclwb01 PROC /usr/sbin/httpd 12 20 PROC "Xvnccluster :99" 1 DISK "/prodQ8/services" 20 30 DEFAULT # These are the built-in defaults. UP 1h LOAD 5.0 10.0 DISK * 90 95 MEMPHYS 100 101 MEMSWAP 50 80 MEMACT 90 97 Getting the configreport of one of the itromclwb01 i get some strange thing (it's possible that those are correct...): I don't get the limits for the filesystems, i get only the default and the procs list is somewhat unusable/unreadable... disk No -/-/- Default limits: Yellow 90% full, Red 95% full /prodQ8/services /prodQ8/qmail /boot /prodQ8/cups /root /prodQ8/squid .....

1 0