[Xymon] xymon disk not alerting at 100%, need another set of eyes

Thomas Eckert thomas.eckert at it-eckert.de
Fri Jan 6 08:49:02 CET 2017


[ kind of a “me too” note ]

This most likely explains a situation I observed with a busybox-based system (QNap) that also graphed all values just fine but happily showed green despite 100% full.
I was expecting the graphing and alerting/color-determination to use the same parser … which is obviously not the case.

Thanks for the investigation.
> On 05 Jan 2017, at 22:46, Scot Kreienkamp <Scot.Kreienkamp at la-z-boy.com> wrote:
> 
> The –P was it, strange that it would still receive and graph the value but not be able to read it for the testing piece.  I only have the –h in there because I always do that by default without thinking about it, it’s not in the script that xymon runs though.  
>  
> Thank you, Paul and JC!  
>  
> Scot Kreienkamp  | Senior Systems Engineer | La-Z-Boy Corporate
> One La-Z-Boy Drive | Monroe, Michigan 48162 | Office: 734-384-6403 | | Mobile: 7349151444 | Email: Scot.Kreienkamp at la-z-boy.com <mailto:Scot.Kreienkamp at la-z-boy.com>
> From: Root, Paul T [mailto:Paul.Root at CenturyLink.com] 
> Sent: Thursday, January 5, 2017 4:41 PM
> To: Root, Paul T; Scot Kreienkamp; 'Japheth Cleaver'; 'shea4th at comcast.net'
> Cc: 'xymon'
> Subject: RE: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Specifically if you look in the xymonclient-linux shell script, the df output is looking for:
>  
> EXCLUDES=`cat /proc/filesystems | grep nodev | grep -v rootfs | awk '{print $2}' | xargs echo | sed -e 's! ! -x !g'`
> ROOTFS=`readlink -m /dev/root`
> df -Pl -x iso9660 -x $EXCLUDES | sed -e '/^[^   ][^     ]*$/{
> N
> s/[     ]*\n[   ]*/ /
> }' -e "s&^rootfs&${ROOTFS}&"
> echo "[inode]"
> df -Pil -x iso9660 -x $EXCLUDES | sed -e '/^[^  ][^     ]*$/{
> N
> s/[     ]*\n[   ]*/ /
> }' -e "s&^rootfs&${ROOTFS}&"
>  
> So it specifically does not want –h.     That is most likely the problem.
>  
> From: Root, Paul T 
> Sent: Thursday, January 05, 2017 3:35 PM
> To: 'Scot Kreienkamp'; Japheth Cleaver; shea4th at comcast.net <mailto:shea4th at comcast.net>
> Cc: xymon
> Subject: RE: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> I think you need df –P in your sudo script.
>  
> From: Xymon [mailto:xymon-bounces at xymon.com <mailto:xymon-bounces at xymon.com>] On Behalf Of Scot Kreienkamp
> Sent: Thursday, January 05, 2017 3:32 PM
> To: Japheth Cleaver; shea4th at comcast.net <mailto:shea4th at comcast.net>
> Cc: xymon
> Subject: Re: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Here’s the output of df, just looks normal to me:
>  
> [root at corpvskreienl bin]# df -h
> Filesystem           Size  Used Avail Use% Mounted on
> /dev/mapper/ol-root  100G  3.2G   97G   4% /
> devtmpfs             909M     0  909M   0% /dev
> tmpfs                920M   72K  920M   1% /dev/shm
> tmpfs                920M   49M  872M   6% /run
> tmpfs                920M     0  920M   0% /sys/fs/cgroup
> /dev/sda1            2.0G  2.0G   20K 100% /boot
> tmpfs                184M     0  184M   0% /run/user/0
>  
> <image001.png>
> From: Japheth Cleaver [mailto:cleaver at terabithia.org <mailto:cleaver at terabithia.org>] 
> Sent: Thursday, January 5, 2017 3:55 PM
> To: Scot Kreienkamp; shea4th at comcast.net <mailto:shea4th at comcast.net>
> Cc: xymon
> Subject: Re: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Hmm. That seems strange... We're not showing any space data there, just 0%. It should look something like this below... 
> What's the output of the normal 'df' command on these boxes?
> 
> 7416 2017-01-05 12:49:20.809413 Disk check host rhel5-i386.build
> 7416 2017-01-05 12:49:20.809926 Disk check: header 'Filesystem         1024-blocks      Used Available Capacity Mounted on', columns 3 and 4
> 7416 2017-01-05 12:49:20.810409 Disk check: FS='/' level 74%/1751688U (thresholds: 90/95, abs: 0/0)
> 7416 2017-01-05 12:49:20.810818 Disk check: FS='/boot' level 13%/83419U (thresholds: 90/95, abs: 0/0)
> 7416 2017-01-05 12:49:20.820510 Disk check: FS='/dev/shm' level 1%/257372U (thresholds: 90/95, abs: 0/0)
> 7416 2017-01-05 12:49:20.820691 Adding to combo msg: status rhel5-i386,build.disk green Thu Jan  5 12:49:52 PST 2017 - Filesystems ok
> 
> One idea: Are these the same boxes that you had to put the sudo hack in for? Is it possible the arguments to 'df' are not being passed in with the execution? At the very least, I think missing a -P (posix) could cause parsing problems.
> 
> -jc
> 
> On 1/5/2017 12:23 PM, Scot Kreienkamp wrote:
> Running the config dump with xymoncmd in front of it didn’t make any difference to the output. 
>  
> Here’s the debug mode output for the disk section:
>  
>  
> 4873 2017-01-05 15:18:07.632660 Disk check host corpvskreienl.na.lzb.hq
> 4873 2017-01-05 15:18:07.632670 Disk check: header 'Filesystem          1K-blocks    Used Available Use% Mounted on', columns 3 and -1
> 4873 2017-01-05 15:18:07.632677 Disk check: FS='/' level 0%/101469992U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632683 Disk check: FS='/dev' level 0%/930248U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632688 Disk check: FS='/dev/shm' level 0%/941992U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632692 Disk check: FS='/run' level 0%/892380U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632696 Disk check: FS='/sys/fs/cgroup' level 0%/942064U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632700 Disk check: FS='/boot' level 0%/20U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632704 Disk check: FS='/run/user/0' level 0%/188416U (thresholds: 90/95, abs: 0/0)
> 4873 2017-01-05 15:18:07.632708 Adding to combo msg: status corpvskreienl,na,lzb,hq.disk green Thu Jan  5 15:18:07 EST 2017 - Filesystems ok
> 4873 2017-01-05 15:18:07.632710 combo_add (tcp): current xymonmsg size: 11068, buffer size: 617; maxmsgspercombo: 100, messages queued so far: 2
> 4873 2017-01-05 15:18:07.632713 Inode check host corpvskreienl.na.lzb.hq
> 4873 2017-01-05 15:18:07.632719 Inode check: header 'Filesystem          1K-blocks    Used Available Use% Mounted on', columns -1 and -1
> 4873 2017-01-05 15:18:07.632726 Inode check: FS='/' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632752 Inode check: FS='/dev' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632758 Inode check: FS='/dev/shm' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632762 Inode check: FS='/run' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632766 Inode check: FS='/sys/fs/cgroup' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632770 Inode check: FS='/boot' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632773 Inode check: FS='/run/user/0' level 0%/0U (thresholds: 70/90, abs: 0/0)
> 4873 2017-01-05 15:18:07.632777 Adding to combo msg: status corpvskreienl,na,lzb,hq.inode green Thu Jan  5 15:18:07 EST 2017 - Filesystems ok
>  
>  
>  
>  
> From: Japheth Cleaver [mailto:cleaver at terabithia.org <mailto:cleaver at terabithia.org>] 
> Sent: Thursday, January 5, 2017 3:11 PM
> To: Scot Kreienkamp; shea4th at comcast.net <mailto:shea4th at comcast.net>
> Cc: xymon
> Subject: Re: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Eyeballing it, it seems to be correct, and if windows matches are working then it seems like class (or at least OS) is being sensed properly. Can you put xymond_client in debug mode (-USR2) and show the output from the processing of the disk section for this client? It should indicate there the thresholds it *thinks* apply to this host.
> 
> Also, when running manually like this:
>     /usr/libexec/xymon/xymond_client --dump-config
> ...can you prefix with xymoncmd and see if anything changes? Weird configs I'd forgotten about in xymonserver.cfg have bit me on occasion.
> 
> -jc
> 
> On 1/5/2017 11:38 AM, Scot Kreienkamp wrote:
> No… it’s showing up on the page and in the graph.  Even if it was ignored, reverting to the default out-of-the-box config would have removed the ignore also. 
>  
> <image002.jpg>
> From:shea4th at comcast.net <mailto:shea4th at comcast.net>[mailto:shea4th at comcast.net <mailto:shea4th at comcast.net>] 
> Sent: Thursday, January 5, 2017 2:36 PM
> To: Scot Kreienkamp
> Cc:cleaver at terabithia.org <mailto:cleaver at terabithia.org>; xymon
> Subject: Re: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Is /boot ignored?
>  
>  
> It’s not the partition the client is on, and it’s been that way for days. 
>  
> So a bit more troubleshooting, I moved all the files out of analysis.d so the only analysis config is the default included from the install and restarted xymon. 
>  
> [root at monvxymon analysis.d]# /usr/libexec/xymon/xymond_client --dump-config --config=etc/analysis.cfg   ; echo Done
> UP 3600 -1 (line: 365)
> LOAD 5.00 10.00 (line: 366)
> DISK * 90% 95% 0 -1 red (line: 367)
> INODE * 70% 90% 0 -1 red (line: 368)
> MEMREAL 100 101 (line: 369)
> MEMSWAP 50 80 (line: 370)
> MEMACT 90 97 (line: 371)
> Done
>  
>  
> Then I restarted my client to force it to report in.  The disk test is still green with the /boot partition at 100% full!  All my windows clients are working, but NONE of my Linux clients with disk full conditions are working.  
>  
> Something is definitely broken!
>  
> JC, any ideas? 
>  
> From:shea4th at comcast.net <mailto:shea4th at comcast.net>[mailto:shea4th at comcast.net <mailto:shea4th at comcast.net>] 
> Sent: Thursday, January 5, 2017 2:18 PM
> To: Scot Kreienkamp
> Cc: xymon
> Subject: Re: [Xymon] xymon disk not alerting at 100%, need another set of eyes
>  
> Hi Scott,
>  
> What may have happened is that the disk filled up quicker than the client could send the alert.
> If the client is on the same disk that is full.  That's caught me a few times.
>  
> HTH
> Regards
> Greg Shea
>  
> So I had another thought, I copied the class statement to another file so it’s now first in the list and last in the list, and my disk test is still green.  Is the class match broken?
>  
> I’m on 4.3.27-1 from Terabithia.
>  
> Thanks!
>  
> From: Scot Kreienkamp 
> Sent: Thursday, January 5, 2017 1:53 PM
> To:xymon at xymon.com <mailto:xymon at xymon.com>
> Subject: RE: xymon disk not alerting at 100%, need another set of eyes
>  
> After re-reading I can see how that may not be totally clear.  By alerting, I mean that the disk test is still green, even though a partition is at 100%full.  
>  
>  
>  
>  
> I found two hosts that weren’t alerting on disk full condition and started digging into the problem further.  As I understand it, xymon matches the first entry from analysis config files.  So I dumped the analysis config for disks:
>  
> Client line:
> [collector:]
> client corpvskreienl,na,lzb,hq.linux linux
>  
> [root at monvxymon hosts.d]# /usr/libexec/xymon/xymond_client --dump-config --config=etc/analysis.cfg |grep -i ^disk
> DISK %^(D|E|F|G|H|I|J|K|L|M|N|O|P|Q|R|S|T|U|V|W|X|Y|Z) 15728640U 10485760U 0 -1 red HOST=%(mondbexec.*|mondb.*|retmaildb.*).na.lzb.hq (line: 515)
> DISK %^(1|2|3|4|5|6|7|8|9|0).* IGNORE HOST=%(mondbexec.*|mondb.*|retmaildb.*).na.lzb.hq (line: 516)
> DISK %^(D|E|F|G|H|I|J|K|L|M|N|O|P|Q|R|S|T|U|V|W|X|Y|Z) 15728640U 10485760U 0 -1 red HOST=%(mon|new|red|neo|taz|sil|kin|sal|hpt)exch.*.na.lzb.hq (line: 527)
> DISK %^(1|2|3|4|5|6|7|8|9|0).* IGNORE HOST=%(mon|new|red|neo|taz|sil|kin|sal|hpt)exch.*.na.lzb.hq (line: 528)
> DISK %^(D|E|F|G|H|I|J|K|L|M|N|O|P|Q|R|S|U|V|W|X|Y|Z) 15728640U 10485760U 0 -1 red HOST=%dayexch.*.na.lzb.hq (line: 539)
> DISK %^T IGNORE HOST=%dayexch.*.na.lzb.hq (line: 540)
> DISK %^(1|2|3|4|5|6|7|8|9|0|).* IGNORE HOST=%dayexch.*.na.lzb.hq (line: 541)
> DISK C 204800U 102400U 0 -1 red HOST=mdas4000.mdmza.dmz.hq (line: 567)
> DISK E 101% 101% 0 -1 red HOST=mdas4000.mdmza.dmz.hq (line: 568)
> DISK F 99% 100% 0 -1 red HOST=mons6000.na.lzb.hq (line: 576)
> DISK %^(D|E|F|G|H|I|J|K|L|M|N|O|P|Q|R|S|T|U|V|W|X|Y|Z) 15728640U 10485760U 0 -1 red PAGE=infrastructure/fileserv (line: 582)
> DISK D 99% 100% 0 -1 red HOST=lzbv5223.na.lzb.hq,lzbv6016.na.lzb.hq (line: 746)
> DISK * 90% 95% 0 -1 red HOST=%dvrvas(0|1)\.mdmza.dmz.hq (line: 762)
> DISK * 90% 95% 0 -1 red CLASS=powershell (line: 1054)
> DISK * 90% 95% 0 -1 red CLASS=win32 (line: 1073)
> DISK * 90% 95% 0 -1 red CLASS=linux (line: 1090)
> DISK * 90% 95% 0 -1 red (line: 1132)
>  
>  
> I can’t find any lines above where the hostname matches, it’s on page Infrastructure/Miscellaneous so none of the page statements match, so it should match on the class.  Or the very last line is the system default which should apply if nothing else.  My server is sitting at 100%full on one partition so it SHOULD be alerting. 
>  
> Thanks for any help. 
>  
> This message is intended only for the individual or entity to which it is addressed.  It may contain privileged, confidential information which is exempt from disclosure under applicable laws.  If you are not the intended recipient, you are strictly prohibited from disseminating or distributing this information (other than to the intended recipient) or copying this information.  If you have received this communication in error, please notify us immediately by e-mail or by telephone at the above number. Thank you.
> 
> 
> _______________________________________________
> Xymon mailing list
> Xymon at xymon.com <mailto:Xymon at xymon.com>
> http://lists.xymon.com/mailman/listinfo/xymon <http://lists.xymon.com/mailman/listinfo/xymon>
>  
>  
>  
>  
> 
> This communication is the property of CenturyLink and may contain confidential or privileged information. Unauthorized use of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender by reply e-mail and destroy all copies of the communication and any attachments.
> _______________________________________________
> Xymon mailing list
> Xymon at xymon.com
> http://lists.xymon.com/mailman/listinfo/xymon

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.xymon.com/pipermail/xymon/attachments/20170106/493c43c7/attachment.html>


More information about the Xymon mailing list