January 2009 - Xymon - Xymon mailman web

clientupdate related bug?! need a debug switch for clientupdate
by vadud3＠gmail.com 13 Jan '09

13 Jan '09

On Mon, Jan 5, 2009 at 4:32 PM, Asif Iqbal <vadud3 at gmail.com> wrote: > On Wed, Dec 31, 2008 at 3:24 PM, Asif Iqbal <vadud3 at gmail.com> wrote: >> Hi >> >> I have a list of hosts with CLASS:myclass and I have a tar file >> newimage.tar file in ~hobbit/server/download dir >> >> In my ~hobbit/server/etc/client-local.cfg file I added this entry >> >> [myclass] >> clientversion:newimage >> >> This image suppose to add a new service called `mytest` in the status >> column for all hosts in that particular group >> >> I noticed only few hosts got that new service column `mytest' , after >> a little wait, and most did not create the new column >> I have a total of 40 hosts in that group >> >> I also noticed all of the hosts have the [clientversion] section in >> the client data showing the `newimage' as the version name. >> So hobbit server at least instructed the client correctly to pickup >> the new image and the clients were able to run the clientupdate >> command. >> >> All the clients have the new ext/newscript.sh and the >> etc/clientversion.cfg file (with content newimage) >> >> However, the clients that did not show the new service column is >> _still_ carrying have the old etc/clientlaunch.cfg file. It does not >> have an entry of the [newservice] section. So that explains why new >> service column is not showing in status page for those hosts. >> >> A working and non-working clients' hobbitclient.log file has same >> 'tar: blocksize = 0' . So tar xf probably did not fail. >> They all are same OS, solaris 8. So something in the non-working hosts >> preventing it from overwriting the old etc/clientlaunch.cfg file >> >> How do I debug this? > > I made another update and same issue. Some of the hosts did not get > the new clientlaunch.cfg file. > I had to manually go through each of them just to add the snippet > related to the new extension script. > > Would be nice if I could avoid that, afterall `clientupdate' claims to > do just that I ran the following on two hosts, one where clientupdate works and one where it does not. I did not see it ever touched the clientlaunch.cfg file for either of them. I am probably not using the correct switches with truss. (hobbit):~/client$ rm etc/clientversion.cfg (hobbit)myhost:~/client$ export HOBBITCLIENTHOME=`pwd` (hobbit)myhost:~/client$ truss -f ./bin/bbcmd --env=etc/hobbitclient.cfg ./bin/clientupdate --update=newversion This really begs a debug switch for clientupdate > > >> >> >> >> -- >> Asif Iqbal >> PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu >> > > > > -- > Asif Iqbal > PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu > -- Asif Iqbal PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu

1 2

hobbitlaunch: error due to invocation timing
by nelis＠nlr.nl 13 Jan '09

13 Jan '09

Hello all, one test in Xymon, actually Hobbit 4.3.0, is performed hourly, thus in hobbitlaunch.cfg "INTERVAL 60m" is specified. The script delivers two totals of the last hour, which are collected in an RRA, type is GAUGE, and are displayed in Xymon. At first glance it works great. However, the problem is that the displayed values (and thus the values stored in the RRA) differ from the measured values, and the difference is growing. It turns out that the difference is caused by interpolation of RRD: it computes the value at the last time-step using linear interpolation. Additionally, the mean time between two successive invocations of the script is 3602 seconds. This implies that the correction caused by interpolation also grows. If Xymon reports a time of test equal to the last time-step, RRD will not perform any interpolation. Is this possible? Alternatively, if hobbitlaunch can start the script at a given time within an hour, like crontab, the difference due to interpolation can be minimized and will be stable. Is this possible? Or is there a better way to have the measured values stored in the RRA? TIA, Wim Nelis. ******************************************************************************************************* The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages. *******************************************************************************************************

1 0

ESX Hosts
by jdustin＠usm.maine.edu 13 Jan '09

13 Jan '09

>>> On 1/12/2009 at 10:38 PM, in message <496C491C.0B6 : 45 : 32950>, "Everett, Vernon" <Vernon.Everett at woodside.com.au> wrote: > Hi all > > Has anybody had any success monitoring ESX hosts yet? > In particular, I am looking for HBA and disk stats, CPU, memory, etc. > The usual suspects. I decided to use a "standard" Hobbit client running on the service console, then add my own customizations: - use vdf for disk space monitoring instead of df (picks up VMDK volumes) - use esxtop to get CPU usage for entire ESX box Here is a snippet I wrote for parsing ESXTOP output: my $esxtop = "sudo /usr/bin/esxtop -b -d $interval"; open(CMD,"$esxtop |") || die "error - could not open $esxtop $!\n"; my $count = 0; while( <CMD> ) { chomp; s/\"//g; my (@row) = split/\,/; $count++; if ( $count <= 2 ) { next; } # skip 1st few results, just titles and 100% values my $util = int($row[9]); # round to INT, because RRD does not like floating-point &SendHobbit($util); if ( $verbose ) { print "util = $util\n"; } } I hope this helps, please don't flame my crappy code... Enjoy! -- Jon Dustin - Network Specialist University of Southern Maine Portland, ME 207-780-4152

2 1

Update hobbit to Xymon
by Samuel.RENARD＠cg29.fr 13 Jan '09

13 Jan '09

Ho all, I've a hobbit server (4.2.0) with about 100 servers. I want to update to xymon, safely... Of course. I've a lot of windows machines and I want to profit of the central mode. How can I update my hobbit server ? There were some mails which talk about, but there were no clear responses. Regards, Samuel Ce message a ete scanne par l'anti-virus du Conseil General du Finistere.

4 4

bbtest-net SSL support
by pmeech＠gmail.com 13 Jan '09

13 Jan '09

Hi All, I've been using hobbit for some time now, and have only just recently found the need to monitor SSL websites. So I've read through the mailing list regarding all the HTTP error 0 and SSL Support posts. Unfortunately I still cannot get bbtest-net --version to report an SSL version, output is just: bbtest-net version 4.2.0. I have OpenSSL installed on my system (Ubuntu 8.04): # openssl version OpenSSL 0.9.8e 23 Feb 2007 ...and when I run configure it finds the SSL libraries: Checking for OpenSSL ... Found OpenSSL include files in /usr/include Found OpenSSL libraries in /usr/lib After running make install (where I can see references to the include directories) and checking in the bbnet folder, my copy of bbtest-net has not been updated (still shows the date I originally compiled hobbit). Does anyone have any suggestions? Many Thanks, Phil

2 3

get hobbit to show status for same host but with different configuration
by mogens.busk＠gmail.com 13 Jan '09

13 Jan '09

Hi I Would like to make Xymon display 2 pages with the same hosts on but with different configuration (etc/hobbit-clients.cfg). Is there a way to make hobbit read another etc/hobbit-clients.cfg eg. etc/hobbit-clients_servicelinea.cfg for one page set and the normal etc/hobbit-clients.cfg for the normal page set? alternatively is it possible to make a virtual host with bbcombotest which includes some specifik process, ports and services from different hosts? -- mvh mogens busk mob at rcf.dk

2 6

[hobbit] vmstat problem
by doctor＠makelofine.org 13 Jan '09

13 Jan '09

Hi all, I'm using Xymon 4.2.2 on both clients and server side. All my OS are linux (almost all are kernel 2.6) My case : One of my client does not send [vmstat] section to the server. After researching a lot, i found that executing manually hobbitclient-linux.sh is working great. I've the hobbit_vmstat.HOSTNAME.$$ file created, but it's not moved to hobbit_vmstat.HOSTNAME. So I modified hobbitclient-linux.sh in order to get directly the hobbit_vmstat.HOSTNAME file, and it works when i start hobbitclient-linux.sh manually. But when i start xymon client, the hobbit_vmstat.HOSTNAME is not generated, and the message sent to my server still doesn't include [vmstat] section I'm sorry if my explanations/vocabulary sucks. Any help would be very appreciated. I've got a second issue with vmstat. On my others clients, I get many (about 10) hobbit_vmstat.$MACHINE.$$ files in $BBTMP dir. Is it a normal behaviour ? Regards, Damien

1 0

two hobbit servers
by denis＠crewhub.net 12 Jan '09

12 Jan '09

hi, Is there a sort of how-to or a guide to configure two hobbit server that monitor each other ? I encoutered problems with classic configuration so i disconnected crossed monitoring between them. kind regards, Denis ROY

3 3

hobbitfetch problem on trunk
by obeau79＠gmail.com 12 Jan '09

12 Jan '09

An HTML attachment was scrubbed... URL: <http://lists.xymon.com/pipermail/xymon/attachments/20090112/21bf7174/attach…>

2 1

RRD log errors
by rmcgraw＠purdue.edu 12 Jan '09

12 Jan '09

Every five minutes and I am getting the following in the rrd-status.log. Host peano is: Hostname: peano.math.purdue.edu OS: Red Hat Enterprise Linux AS release 4 (Nahant Update 6) Linux peano.math.purdue.edu 2.6.9-67.0.1.ELsmp x86_64 I searched the archive and did not find any information about this. My graphs look ok. Any ideas? Robert 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/la.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/procs.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/users.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/clock.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/disk,root.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/disk,boot.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/disk,var.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/memory.real.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/memory.swap.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) 2008-06-04 13:20:59 RRD error updating /local/hobbit420/data/rrd/peano.math.purdue.edu/memory.actual.rrd from 128.210.3.244: illegal attempt to update using time 1212600059 when last update time is 1212600059 (minimum one second step) -------------------------------------------------------------------- Robert P. McGraw, Jr. Manager, Computer System EMAIL: rmcgraw at purdue.edu Purdue University ROOM: MATH-807 Department of Mathematics PHONE: (765) 494-6055 150 N. University Street West Lafayette, IN 47907-2067

6 6