July 2006 - Xymon - Xymon mailman web

[hobbit] Proxy
by brodie＠mcw.edu 14 Jul '06

14 Jul '06

Also, what IP address would I use on the hobbit server for those hosts being monitored? (that is, how is the "conn" test configured?) -----Original Message----- From: Henrik Stoerner [mailto:henrik at hswn.dk] Sent: Friday, July 14, 2006 3:34 PM To: hobbit at hswn.dk Subject: Re: [hobbit] Proxy On Fri, Jul 14, 2006 at 02:18:58PM -0500, Brodie, Kent wrote: > Hi- can someone briefly explain how the bbproxy stuff works? We > have a cluster whose "head node" is accessible, but the cluster nodes > are hidden behind that server. > > If someone can explain how something like this is set up, that'd be > super.. > > ============================================================= (network) > [hobbit server] [most of my nodes] [head node] > || > || > [cluster nodes] Since your cluster nodes can talk to the head node, and the head node can talk to the hobbit server, the most simple solution would be to run bbproxy on the head node of the cluster. Your cluster nodes would then use the head node's bbproxy as their BBDISPLAY server. bbproxy on the head node would be configured to forward messages to the real hobbit server. Henrik To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

1 0

Bug Suspect - Hobbit 4.2-beta-20060605
by M.N.Harron＠herts.ac.uk 14 Jul '06

14 Jul '06

Hi, I'm not sure if this message is going to a list or an individual. Please accept my apology if this problem has been reported previously. The script $HOME/client/runclient.sh is created without a space character after the "sed -e" statements. This causes the interpretation of options "--hostname", "--class", etc to fail. So, for example, the line: MACHINEDOTS="`echo $1 | sed -e's/--hostname=//'`" should read: MACHINEDOTS="`echo $1 | sed -e 's/--hostname=//'`" ^ I have confirmed this problem on Solaris 9, Solaris 10, Red Hat 9 and Fedora Core 5 so I guess that it is not limited to any particular platform. Best regards, Mike ====================================================================== | Mike Harron | Network Development and | Security Manager | E-mail: M.N.Harron at herts.ac.uk | Learning Resources Centre | Telephone (external): 01707 286 447 University of Hertfordshire | Telephone (internal): 3447 College Lane | Hatfield | Telephone (mobile): 07984 605 275 AL10 9AB | ======================================================================

2 1

[hobbit] A funny thing happened on the way to the RC...............
by brodie＠mcw.edu 14 Jul '06

14 Jul '06

Henrik- no trouble at all. That's why it's a release CANDIDATE-- and why we're willing to put up with issues like this for the sake of making a REALLY great product. There's no way you would find all these little issues without people like us willing to play around and put it through its paces. And yeah, it got my attention. And my boss's. And my sysadmin. And the pager company who stopped accepting emails after the first 20 or so. That's what we do. :-) --------------------------------------------------------- Kent C. Brodie - brodie at phys.mcw.edu Department of Physiology Medical College of Wisconsin (414) 456-8590 -----Original Message----- From: Henrik Stoerner [mailto:henrik at hswn.dk] Sent: Friday, July 14, 2006 12:37 AM To: hobbit at hswn.dk Subject: Re: [hobbit] A funny thing happened on the way to the RC............... On Thu, Jul 13, 2006 at 04:48:08PM -0500, Brodie, Kent wrote: > Fired up BB. Then, everyone's pager went NUTS. 100 pages, times 4 > pagers. WHOOOOPS. The entire world had uh..... "lost > connectivity." > > What had really happened of course is, that the hobbit userid couldn't > run fping, since I had forgotten to suid the executable. And yes, I > did eventually find it in a log file, but... But, this had the > very nasty side effect of turning each and every connectivity icon RED. > And then the alerts kicked in..... And four Duracell batteries died. Well, at least Hobbit managed to get your attention that there was some configuration problem :-) Seriously - you're right, it shouldn't do that. Such a problem should end up with a red status in the "bbtest" column and nothing else. I'll make sure it does that. Sorry about the trouble! Regards, Henrik To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

1 0

bb-xsnmp problem
by damiano.vozza＠apss.tn.it 14 Jul '06

14 Jul '06

I use Hobbit 4.1.2.p1 on FC4 and I tried to test bb-xsnmp.pl v. 1.79 with a Cisco router. I ran the test on command line with /usr/local/hobbit/server/bin/bbcmd bb-xsnmp.pl and all worked fine. I tried to launch it from hobbitlaunch [xsnmp] ENVFILE /usr/local/hobbit/server/etc/hobbitserver.cfg NEEDS hobbitd CMD /usr/local/hobbit/server/ext/bb-xsnmp.pl LOGFILE $BBSERVERLOGS/bb-xsnmp.log INTERVAL 5m but hobbitlaunch process used a lot of VM and the machine slows down; I had to stop Hobbit . No bb-xsnmp.log was produced. Some help? Damiano Vozza Azienda Provinciale per i Servizi Sanitari Servizio Sistemi Informativi Ospedale S. Chiara Via Crosina Sartori 6 Trento - Italy tel. (+39) 0461364856 fax. (+39) 0461903593 e-mail damiano.vozza at apss.tn.it <mailto:damiano.vozza at apss.tn.it> web site www.apss.tn.it <http://www.apss.tn.it/>

1 0

[hobbit] Error in hobbitdboard request crashes hobbitd
by tschenk＠pironet-ndh.com 14 Jul '06

14 Jul '06

> > > > I am testing the 4.2-RC on Linux 2.6.15. Trying all the new features > > (thanks, henrik!) I sent a hobbitdboard request with an error in it: > > > > bb $BBDISP "hobbitdboard status=yellow" > > (should have been color=yellow) > > > > This causes the hobbitd process to crash with a segmentation fault. > > Thanks, this is indeed a bug. A patch is available on the betapatches > page at http://www.hswn.dk/hobbitsw/betapatches/ > Applied the patch, works good now. Thanks, Thomas

1 0

Error in hobbitdboard request crashes hobbitd
by tschenk＠pironet-ndh.com 14 Jul '06

14 Jul '06

Hello, I am testing the 4.2-RC on Linux 2.6.15. Trying all the new features (thanks, henrik!) I sent a hobbitdboard request with an error in it: bb $BBDISP "hobbitdboard status=yellow" (should have been color=yellow) This causes the hobbitd process to crash with a segmentation fault. This is the gdb output of the core file: #0 0xb7ea983b in raise () from /lib/tls/libc.so.6 #1 0xb7eaafa2 in abort () from /lib/tls/libc.so.6 #2 0x0805baf1 in sigsegv_handler (signum=0) at sig.c:57 #3 <signal handler called> #4 0xb7ef9363 in strlen () from /lib/tls/libc.so.6 #5 0x0804e496 in setup_filter (buf=0x8495cd0 "hobbitdboard", defaultfields=0x805fde0 "hostname,testname,color,flags,lastchange,logtime,validtime,acktime,disa bletime,sender,cookie,line1", spage=0xbffdb404, shost=0xbffdb408, snet=0xbffdb40c, stest=0xbffdb410, scolor=0xbffdb414, acklevel=0xbffdb418, fields=0xbffdb41c, chspage=0xbffdb420, chshost=0xbffdb424, chsnet=0xbffdb428, chstest=0xbffdb42c) at hobbitd.c:2066 #6 0x0805117c in do_message (msg=0x82d86a0, origin=0x805e363 "") at hobbitd.c:2753 #7 0x08054478 in main (argc=7, argv=0xbffdf004) at hobbitd.c:4340 Sending this request from any bb or hobbit client seems to crash hobbitd. Regards, Thomas

2 1

A funny thing happened on the way to the RC...............
by brodie＠mcw.edu 14 Jul '06

14 Jul '06

So, I updated my server to the RC version today. Saved all of my configuration files, and (based on info from this list!), decided to add fping to my system this time. Fping installed, no problem. Fired up BB. Then, everyone's pager went NUTS. 100 pages, times 4 pagers. WHOOOOPS. The entire world had uh..... "lost connectivity." What had really happened of course is, that the hobbit userid couldn't run fping, since I had forgotten to suid the executable. And yes, I did eventually find it in a log file, but... But, this had the very nasty side effect of turning each and every connectivity icon RED. And then the alerts kicked in..... And four Duracell batteries died. I don't know how easy it is to implement, but I have a suggestion worth looking into? If fping "doesn't work" (as opposed to, "cannot ping a host"), the result status should be CLEAR, not RED. "test not functional" or "test disabled" or something. Just a thought!!!! :-) --------------------------------------------------------- Kent C. Brodie - brodie at phys.mcw.edu Department of Physiology Medical College of Wisconsin (414) 456-8590

2 1

[hobbit] RC client release bug?
by David.Gore＠verizonbusiness.com 14 Jul '06

14 Jul '06

Hello, We have seen this with recent snapshots and the latest release candidate client. logfetch hangs which causes the client to hang and go purple for all the tests. It can be resolved by killing logfetch and deleting all the entries in ~/client/tmp. We could try to be more surgical on the deleting of files. This has happened on two very independent hosts running Solaris 8, one being a SunFire 880 and another being an E4500/E5500. Suggestions? It can run for many days before hanging. -- David

2 6

Multiple User Admin mode
by steve.aiello＠ge.com 13 Jul '06

13 Jul '06

Today I am using BigBrother/BBgen 3.6. One of the custom features I have hacked into our system is the ability to have multiple admin users that can put servers into maintenance mode. In our BB configuration we made extensive use of the the HostGroup function. Basically creating an aliasname for a list of devices. This came in handy when writing alert rules, I could do one alert rule for a series of devices. Now for maintenance mode, I hacked the maint.pl script found on deadcat. That script was then put into a directory that apache password protected. When you logged in, the maint.pl script compared your login name to any HostGroup that was defined, and if the login name matched a HostGroup the maintenace device listing only listed the devices that were defined by the the HostGroup. Now I did see the archived mailing list article, Re: [hobbit] New Features #2 (the 'publish-cookie' option) (http://www.hswn.dk/hobbiton/2005/05/msg00369.html) That solution really only limits users from viewing different pages. The maint option under the device info report is located under the /hobbit-cgi/ and I do not see a way to limit users from disabling devices they shouldn't. If a user has an admin password, they seem to have admin rights to all of hobbit. Under Hobbit I do not see anyway to define HostGroups, so my old method will not work. Using HostGroupings seemed ideal in that it linked alerts being sent to users and the user's ability to disable devices/tests. ~Steve

2 1

Migration approach
by jeffnewman75＠gmail.com 13 Jul '06

13 Jul '06

All, I was thinking now that the new version of hobbit has moved on to RC (yay!) I would migrate my existing production hobbit server to the new version. What sort of migration approach is recommended? I ask this because I was thinking that it might be better to start fresh with the RRD's because the new version adds some that didn't exist before, and might do things differently with some of the existing ones, etc etc etc.. However, I don't want to lose the historical data, MEANING, I don't have the requirement to re-add the existing data, but would like to be able to refer to it. For example: The new hobbit is running for a week, and someone has XYZ problem, wants me to look at the graphs. I see there is some huge spike at xyz time, but I don't know how long it has been going on, or if this is a growing trend, or whatever, so I want to look at the "old hobbit" graphs for that host which has much more data. Is there any way to "trick" hobbit into looking at the old RRD files for a given host, or is there some approach I am missing? Am I just being dumb? <grin> Comments apprieciated. Thanks, jeff

2 1