- Xymon - Xymon mailman web

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

[hobbit] process checks and user IDs
by gmbfly98＠gmail.com 25 Jul '06

25 Jul '06

I've looked through all the documentation, and can't find any information on how Hobbit processes the process check rules. I ask because we have some BB process check rules that check for user IDs in the ps output, rather than an actual process, and there's no real way of converting that into Hobbit. Also, the Solaris ps output doesn't include the full process name, so I can't check for the process that way either. Any suggestions?

2 1