July 2006 - Xymon - Xymon mailman web

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

INCLUDING annfs volume?
by brodie＠mcw.edu 25 Jul '06

25 Jul '06

I have a server with an NFS mount to a network storage array that I'd like to include in the "disk" test(s). From what I see, the default hobbit behavior si to check local disks only. Is there a clean or proper way to modify the behavior for one client to include an NFS (non-local) disk? --------------------------------------------------------- Kent C. Brodie - brodie at phys.mcw.edu Department of Physiology Medical College of Wisconsin (414) 456-8590

2 1

How do you define days with "TIME" in the hobbit-alerts.cfg?
by James.Gearing＠analog.com 25 Jul '06

25 Jul '06

Hi all, I'm trying to setup my pager to alert me Monday - Friday, 7:00 - 24:00. I understand how to set the hours, but how do you set the days? All the examples I have seen with days in them, only list one day or *. To do multiple days, do you use "," between the days? Can you use a range character? I have also seen Wednesday listed as W or 3 in the examples, will either work? TIME=1,2,3,4,5:700:2400 <---is that right? Thanks, James

3 4