July 2006 - Xymon - Xymon mailman web

conn test fails / comes back
by stephane.caminade＠ias.u-psud.fr 26 Jul '06

26 Jul '06

Hi, I just upgraded to RC, and noticed a very strange thing. When I use fping OR hobbitping, the connection test goes red, then green, and cycles like that for all hosts. I made sure I had no phantom process on the machine by killing everything owned by the hobbit user, then starting the server from scratch, and all tests are fine, except for the connection test. Any idea where that might come from ? Thanks, Stephane -- _____________________________________________________________________________ Stephane Caminade Administrateur Systèmes et Réseaux \ <Stephane.Caminade at ias.u-psud.fr> Institut d'Astrophysique Spatiale / tel : (33) (1) 69 85 87 03 Batiment 121, Universite Paris XI \ fax : (33) (1) 69 85 86 75 F-91405 ORSAY Cedex / www : http://www.ias.u-psud.fr/ _____________________________________________________________________________

3 5

Critical systems summary?
by brodie＠mcw.edu 26 Jul '06

26 Jul '06

Is there some way to show "critical systems" that are defined? The editor only seems to handle one at a time, which I understand... it would be nifty if you could see the "list" of critical systems and the names of the tests that were picked to be critical. Could that perhaps be a choice in the critical systems VIEW? I..e, instead of seeing "red", or "red+yellow", could we add "red+yellow+green" ?? Phrased differently, while it's cool and nice to see critical systems that need attention, I think it's also handy to "see a lit of what we defined to be critical" ??? Just a thought- thanks! -kcb

1 0

RECOVERED flag in hobbit-alerts.cfg question
by brodie＠mcw.edu 26 Jul '06

26 Jul '06

Hi all! As I expand my alerting ruleset to do cool things like only page me about "printers" during the day and so on, I had a weird occurrence today. Basically, I have rules that list less-critical things first, and what I'm trying to do is something like this: Rules for printers: Send emails, pages when printers go offline, and recover STOP HERE, because I do not need other rules to apply (like "connectivity" below) Rules for "connectivity" for anything: Send emails, pages. What is happening is this: When the device (printer) went offline, I got alerted. Yay! I then got alerted exactly two hours later. Exactly what I want. Yay again! Then, the printer recovered. I got *two* emails and *two* pages, because presumably the printers rule *AND* the "connectivity" rule applied, even though I only want the one rule to apply. Am I missing some intended behavior, or is the "recovered" flag ignoring how I want my rules to "stop" when certain conditions are met? My notifications log: Printer goes offline: ====================== Tue Jul 25 16:50:29 2006 lp4.phys.mcw.edu.conn (141.106.188.244) sysadmins[158] 1153864229 500 Tue Jul 25 16:50:29 2006 lp4.phys.mcw.edu.conn (141.106.188.244) pagers[159] 1153864229 500 Same message sent two hours later(yay!) ======================================== Tue Jul 25 18:50:32 2006 lp4.phys.mcw.edu.conn (141.106.188.244) sysadmins[158] 1153871432 500 Tue Jul 25 18:50:32 2006 lp4.phys.mcw.edu.conn (141.106.188.244) pagers[159] 1153871432 500 *DOUBLE* messages sent when recovered state occurs...??? =============================================== Tue Jul 25 19:49:21 2006 lp4.phys.mcw.edu.conn (141.106.188.244) sysadmins[158] 1153874961 500 10792 Tue Jul 25 19:49:22 2006 lp4.phys.mcw.edu.conn (141.106.188.244) pagers[159] 1153874961 500 10792 Tue Jul 25 19:49:22 2006 lp4.phys.mcw.edu.conn (141.106.188.244) sysadmins[175] 1153874961 500 10792 Tue Jul 25 19:49:22 2006 lp4.phys.mcw.edu.conn (141.106.188.244) pagers[176] 1153874961 500 10792 My ruleset for alerts: # These rules change defaults for printers warnings/alerts (only email or # page every 2 hours) # use of IGNORE rule means NO OTHER RULE matches printers after this rule.... HOST=%^pr.*mcw\.edu MAIL sysadmins REPEAT=120 RECOVERED FORMAT=TEXT MAIL pagers REPEAT=120 RECOVERED FORMAT=SMS IGNORE HOST=%^hp.*mcw\.edu MAIL sysadmins REPEAT=120 RECOVERED FORMAT=TEXT MAIL pagers REPEAT=120 RECOVERED FORMAT=SMS IGNORE HOST=%^lp.*mcw\.edu MAIL sysadmins REPEAT=120 RECOVERED FORMAT=TEXT MAIL pagers REPEAT=120 RECOVERED FORMAT=SMS IGNORE # Anything that loses connectivity, email/page every 30 minutes SERVICE=conn MAIL sysadmins REPEAT=30 RECOVERED FORMAT=TEXT MAIL pagers REPEAT=30 RECOVERED FORMAT=SMS

1 0

hobbitd restarting.
by paul.moore＠verizonbusiness.com 25 Jul '06

25 Jul '06

I'm having a problem running the release version of hobbit restarting periodically. The only thing that I can find in any of the logs is the following. Hobbitd.log 2006-07-25 21:50:00 Fatal error in select: Bad file number This is running on a Solaris2 server. Thanks for any help Paul

1 0

[hobbit] darwin "ps" display parsing in OS-X Tiger causes all procs to vanish in display
by brodie＠mcw.edu 25 Jul '06

25 Jul '06

Um, ok, so the parsing works, but that breaks a browser by attempting to do the <.... I was able to fix it simply by removing the STATE column from the process display for that one server.... --------------------------------------------------------- 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: Tuesday, July 25, 2006 4:03 PM To: hobbit at hswn.dk Subject: Re: [hobbit] darwin "ps" display parsing in OS-X Tiger causes all procs to vanish in display On Tue, Jul 25, 2006 at 03:52:38PM -0500, Brodie, Kent wrote: > > OS-X 10.4 (procs display is broken) > > PID PPID USER STARTED STAT PRI %CPU TIME %MEM RSS > VSZ COMMAND > 1 0 root 13Jul06 S<s 32 0.0 0:17.23 -0.0 500 28348 /sbin/launchd > > It's the first line that's causing the hobbit procs parsing to break. > Specifically, the "S<s" status. While I don't know enough OS-X tobe > dangerous, I suspect it's a new type of process status under the latest > Darwin. The "<" in the status is normal, and this is causing the procs > parsing in hobbit to cut off everything beyond that. > This of course makes any procs check useless (in my case,I was checking > for simple things, like "cron", etc - and the test fails of course, > since there's no such (or any other!) process listed.................. I can see why it would stop it from appearing on the web page (since your browser then interprets everything after the '<' as an unknown - very large - HTML tag. But the client parsing code works regardless of this, and should handle the process listing just fine. Could you verify if the following URL http://www.hswn.dk/hobbit-cgi/bb-hostsvc.sh?CLIENT=voodoo.hswn.dk&sectio n=ps (adapted to match you Hobbit server hostname and the OSX client name, of course) returns the full "ps" listing, or only a part of it? Regards, Henrik To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

2 2

[hobbit] darwin "ps" display parsing in OS-X Tiger causes all procs to vanish in display
by brodie＠mcw.edu 25 Jul '06

25 Jul '06

When I click on that url, I am asked to *download* a file called bb-hostsvc.sh... my browser (IE) won't display anything... --------------------------------------------------------- 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: Tuesday, July 25, 2006 4:03 PM To: hobbit at hswn.dk Subject: Re: [hobbit] darwin "ps" display parsing in OS-X Tiger causes all procs to vanish in display On Tue, Jul 25, 2006 at 03:52:38PM -0500, Brodie, Kent wrote: > > OS-X 10.4 (procs display is broken) > > PID PPID USER STARTED STAT PRI %CPU TIME %MEM RSS > VSZ COMMAND > 1 0 root 13Jul06 S<s 32 0.0 0:17.23 -0.0 500 28348 /sbin/launchd > > It's the first line that's causing the hobbit procs parsing to break. > Specifically, the "S<s" status. While I don't know enough OS-X tobe > dangerous, I suspect it's a new type of process status under the latest > Darwin. The "<" in the status is normal, and this is causing the procs > parsing in hobbit to cut off everything beyond that. > This of course makes any procs check useless (in my case,I was checking > for simple things, like "cron", etc - and the test fails of course, > since there's no such (or any other!) process listed.................. I can see why it would stop it from appearing on the web page (since your browser then interprets everything after the '<' as an unknown - very large - HTML tag. But the client parsing code works regardless of this, and should handle the process listing just fine. Could you verify if the following URL http://www.hswn.dk/hobbit-cgi/bb-hostsvc.sh?CLIENT=voodoo.hswn.dk&sectio n=ps (adapted to match you Hobbit server hostname and the OSX client name, of course) returns the full "ps" listing, or only a part of it? Regards, Henrik To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

2 1

Current Version / Patches
by rgordonjr＠gmail.com 25 Jul '06

25 Jul '06

Hi Henrik - I am planning to update the hobbit server (and clients) at work in a few days. If I use the current RC version (and the all-in-one patch set) will that be esentually the release version or would it be better to wait? Great system BTW... -- --==[ Bob Gordon ]==--

2 1

darwin "ps" display parsing in OS-X Tiger causes all procs to vanish in display
by brodie＠mcw.edu 25 Jul '06

25 Jul '06

Hi guys. OK, enough asking a million questions (thank you all!), now it's time for bug reporting. Systems: Apple X-Servers. I have two. The one running OS-X 10.3 is fine. The one running OS-X Tiger 10.4 is being weird. And I figured it out. The symptom on the newer OS is that under the "procs" column, all I see is the process header line, followed by only a small part of the first process listed- and NOTHING else. It looks like this on the web page: PID PPID USER STARTED STAT PRI %CPU TIME %MEM RSS VSZ COMMAND 1 0 root 13Jul06 S After digging, I figured out what's happening. OS-X 10.3: (normal) PID PPID USER STARTED STAT PRI %CPU TIME %MEM RSS VSZ COMMAND 1 0 root 3Jul06 Ss 31 0.0 0:00.21 -0.0 244 18072 /sbin/init 2 1 root 3Jul06 Ss 31 0.0 4:38.77 -0.0 200 18608 /sbin/mach_init 83 1 root 3Jul06 Ss 31 0.0 0:12.67 -0.0 200 18092 /usr/sbin/syslogd -s -m 0 89 1 root 3Jul06 Ss 31 0.0 0:02.68 -0.0 792 29120 kextd 91 2 root 3Jul06 Ss 31 9.0 80:57.57 -0.0 1040 31328 /usr/sbin/configd 92 2 root 3Jul06 Ss 31 0.0 0:00.87 -0.0 892 27888 /usr/sbin/diskarbitrationd 97 1 root 3Jul06 Ss 31 0.0 1:26.87 -0.0 212 18676 /usr/sbin/notifyd 112 1 daemon 3Jul06 Ss 31 0.0 0:00.17 -0.0 144 18084 portmap ..............etc OS-X 10.4 (procs display is broken) PID PPID USER STARTED STAT PRI %CPU TIME %MEM RSS VSZ COMMAND 1 0 root 13Jul06 S<s 32 0.0 0:17.23 -0.0 500 28348 /sbin/launchd 27 1 root 13Jul06 Ss 63 0.0 0:00.00 -0.0 168 27264 /sbin/dynamic_pager -F /private/var/vm/swapfile 31 1 root 13Jul06 Ss 31 0.0 0:06.04 -0.2 3608 30640 kextd 62 1 root 13Jul06 Ss 31 0.0 54:48.59 -0.1 1208 31572 hwmond 63 1 root 13Jul06 Ss 31 0.0 0:00.04 -0.0 600 27844 /usr/sbin/KernelEventAgent 64 1 root 13Jul06 Ss 31 0.0 0:51.95 -0.1 1096 28064 /usr/sbin/mDNSResponder -launchdaemon 65 1 root 13Jul06 Ss 31 0.0 0:11.21 -0.0 572 27592 /usr/sbin/netinfod -s local 66 1 root 13Jul06 Ss 63 0.0 53:42.36 -0.4 7400 47040 servermgrd -x 67 1 root 13Jul06 Ss 31 0.0 0:25.95 -0.0 512 27284 /usr/sbin/syslogd 68 1 root 13Jul06 Ss 63 0.0 0:10.54 -0.0 732 27684 watchdogtimerd ...........etc It's the first line that's causing the hobbit procs parsing to break. Specifically, the "S<s" status. While I don't know enough OS-X tobe dangerous, I suspect it's a new type of process status under the latest Darwin. The "<" in the status is normal, and this is causing the procs parsing in hobbit to cut off everything beyond that. 1 0 root 13Jul06 S<s 32 0.0 0:17.23 -0.0 500 28348 /sbin/launchd This of course makes any procs check useless (in my case,I was checking for simple things, like "cron", etc - and the test fails of course, since there's no such (or any other!) process listed.................. --------------------------------------------------------- Kent C. Brodie - brodie at phys.mcw.edu Department of Physiology Medical College of Wisconsin (414) 456-8590

2 1

"ps" options for darwin
by brodie＠mcw.edu 25 Jul '06

25 Jul '06

Here's from the man of ps: The ">" or "<" are normal. Additional characters after these, if any, indicate additional state information: + The process is in the foreground process group of its control terminal. < The process has raised CPU scheduling priority. > The process has specified a soft limit on memory require- ments and is currently exceeding that limit; such a process is (necessarily) not swapped. A the process has asked for random page replacement (VA_ANOM, from vadvise(2), for example, lisp(1) in a garbage collect). E The process is trying to exit. L The process has pages locked in core (for example, for raw I/O). N The process has reduced CPU scheduling priority (see setpriority(2)). S The process has asked for FIFO page replacement (VA_SEQL, from vadvise(2), for example, a large image processing program using virtual memory to sequentially address voluminous data). s The process is a session leader. V The process is suspended during a vfork. W The process is swapped out. X The process is being traced or debugged. --------------------------------------------------------- Kent C. Brodie - brodie at phys.mcw.edu Department of Physiology Medical College of Wisconsin (414) 456-8590

1 0

Bug : CGI_ACK_OPTS in hobbitcgi.cfg
by Charles.Jones＠ciscolearning.org 25 Jul '06

25 Jul '06

Henrik, I grabbed todays snapshot and noticed that you made the --no-pin (which is a wonderful feature!) option default in hobbitcgi.cfg. The problem is that the option is added outside of the quotes, so that it doesn't take effect: CGI_ACK_OPTS="--env=/home/hobbit/server/etc/hobbitserver.cfg" --no-pin Moving the end quote to the outside of --no-pin fixes it. -Charles

2 1