- Xymon - Xymon mailman web

dbcheck.pl goes purple for all monitored databases?
by david.gore＠verizon.com 04 Sep '12

04 Sep '12

Before I look at spending a lot of time looking at the dbcheck.pl script, I thought I might ask if anyone else has seen the script hang when one of its queries does not return? Typically, dbcheck.pl -fast will hang on a query sent to a database that is having issues and so all the other monitored databases will eventually go purple as the dbcheck.pl is hung. Anybody else see the same issue and implement any solutions? I am going to look at the perl script and see if I can insert some timeout for any and all queries sent to the databases. ~David

1 0

Question on df processing for xymonclient-linux.sh
by Mark.Deiss＠acs-inc.com 04 Sep '12

04 Sep '12

In looking df processing in the Linux Xymon client module in client/xymonclient-linux.sh: echo "[df]" EXCLUDES=`cat /proc/filesystems | grep nodev | awk '{print $2}' | xargs echo | sed -e 's! ! -x !g'` df -Pl -x iso9660 -x $EXCLUDES | sed -e '/^[^ ][^ ]*$/{ N s/[ ]*\n[ ]*/ / }' A bit concerned that the df line can have a dangling "-x" option if the EXCLUDES variable is not populated. With a dangling "-x", the df command will abort with a syntax error and the df status will probably end up with a green status with no rrd history being populated. Users would not be aware unless they specifically brought up the green disk display or were reviewing disk trends and noticing the gap. The EXCLUDES variable can become null if the xymon client account loses read access to /proc/filesystems, for some reason the Linux flavor quit using the nodev tag (maybe started using NODEV or Nodev after a patch update....) or the xymon client went onto an (older) Linux version that did not use the /proc system. Suggest a revision either using sed or awk to build the EXCLUDES variable and a modification of the df command to eliminate the dangling "-x" option EXCLUDES=`cat /proc/filesystems 2>/dev/null | awk ' $1 ~ /nodev/ { printf("-x %s ", $2) }'` Or EXCLUDES=`cat /proc/filesystems 2>/dev/null | sed -ne ' /nodev/{ s/nodev[ ]*/ -x / H } ${ x s/\n//g p }'` df -Pl -x iso9660 $EXCLUDES 2>/dev/null | sed -e '/^[^ ][^ ]*$/{ N s/[ ]*\n[ ]*/ / }' If the parsing for nondev entries in /proc/filesystems now blows up for whatever reason (EXCLUDES is null), the df command will be capturing more than is desired which I believe would be preferred over missing the important file systems. Would like to use "2>&1" rather than "2>/dev/null" in the df line except any captured error lines may cause some downstream havoc with the rrd database processing.

4 5

monitor failed command line logins
by bb4＠richter-it.net 03 Sep '12

03 Sep '12

Hi XYmonians, since I do not want to reinvent the wheel I wanted to ask if somebody already has some script running that monitors specific log files on Unix/Linux for break in attempts and triggers an alert. Regards Torsten

2 1

dbcheck.pl bug with locks check, comments, suggestion welcome
by david.gore＠verizon.com 31 Aug '12

31 Aug '12

We really like the dbcheck.pl script and its ability to monitor remote databases. Unfortunately, the check caused a number of databases to become unreachable and complain about temp space, oracle errors ORA-01114 and ORA-27063. Perhaps someone with a bit more database experience than I can verify the veracity of this statement used for blocking in the dbcheck.pl script: $selcmd{oracle}{locks}=qq#select l1.id1 "id1", l1.id2 "id2", l2.id1 "id3", l2.id2 "id4", l1.sid "sid1", l2.sid "sid2", l1.type "type1", l2.type "type2", l1.request "request1", l2.request "request2", l1.lmode "mode1", l2.lmode "mode2", l1.ctime "ctime1", l2.ctime "ctime2" from v\$lock l1, v\$lock l2 where l1.BLOCK=1 and l2.request > 0 and l1.id1 =l2.id1 and l2.id2 = l2.id2# ; The last bit of the where clause is what I think is suspect: l1.id1 =l2.id1 and l2.id2 = l2.id2 -------------------^^ ----------------------------^^ SHOULD BE: l1.id1 =l2.id1 and l1.id2 = l2.id2 -------------------^^ ----------------------------^^ Any database experts out there? ~David

1 0

xymon 4.3.10 not working for HP-UX 11.31 !
by tomas.andersson＠sca.com 31 Aug '12

31 Aug '12

Hi there! I have still no Luck with the xymon v 4.3.10 to function on my systems that is: HP-UX hpivm006 B.11.31 U ia64 1185740114 unlimited-user license ia64 hp server Integrity Virtual Machine The folloving logfiles only contain rows of: " Peer not up, flushing message queue" alert.log clientdata.log history.log hostdata.log rrd-data.log rrd-status.log the xymonlaunch.log looks OK: 2012-08-29 13:53:59 Loading tasklist configuration from /opt/xymon/server/etc/tasks.cfg 2012-08-29 13:53:59 Loading hostnames 2012-08-29 13:53:59 Loading saved state 2012-08-29 13:53:59 Setting up network listener on 0.0.0.0:1984 2012-08-29 13:53:59 Setting up signal handlers 2012-08-29 13:53:59 Setting up xymond channels 2012-08-29 13:53:59 Setting up logfiles I now have a dev server for test (a clone of the prod server) Here I have installed both the current running version 4.3.7 and the 4.3.10 I've compiled them both OK. I installed the 4.3.7 version and all started and functioned OK. Then I installed the 4.3.10 and this is not functioning at all. I get no messages from the test clients and eventually the system goes purple on all but the conn, info and trend columns. (even for the server itself) Apparently the "xymond_channel" processes starts to eat CPU and after a while I had to stop the system to prevent it >From using to much CPU (since I have a virtual environment and there was prod servers that was affected) Anybody have a clue of what is wrong or anybody has succeded to run the xymopn 4.3.10 on hp-ux ?? Best Regards, /Tomas Andersson Unix System Admin/Tech/Dev SCA IT Services Tel. +46 31 7460313 Mob. +46 703 610313 This message may contain confidential, proprietary or legally privileged information and is intended for the addressee's use only. Any usage, disclosure, distribution, print or copying of any part of this message is prohibited unless you are the intended recipient. If you receive this message in error, please delete it from any computer and notify the sender. If you suspect that this message may have been altered, please notify the sender. SCA has taken every reasonable precaution to ensure that this e-mail and any attachments to this e-mail has been scanned for viruses. However, SCA does not accept liability for any damage caused by software viruses brought to you by this mail.

3 4

Script launched but no column appears
by benjamin.boudoir＠eeple.fr 31 Aug '12

31 Aug '12

Hi, I've recently add a client perl script on some of my linux servers but the Xymon server does not add the column. The script seems to work correctly because I can see it data with bb- hostsvc.sh. I explicitly add the column name in bb-host but nothing's happen. Any ideas? Regards, -- Benjamin Boudoir meltyNetwork / Eeple Administrateur Système

4 4

Reasons to upgrade
by richard.crane＠haskins.yale.edu 29 Aug '12

29 Aug '12

I'm currently running version 4.2.3 on a small (35 host) network, mostly Linux, with a few Mac OS X and one Windows host. What am I missing out on by not upgrading? I am not experiencing any problems (that I know about!) Richard Crane ------- Haskins Laboratories / (203) 865-6163 X 275 / FAX (203) 865-8963 300 George Street/New Haven, CT 06511

3 2

https intermediate certificate check
by bferrell＠baywinds.org 29 Aug '12

29 Aug '12

Hi all, I just did a quick search to see if xymon has a way to check intermediate certificates. The answer *seemed* to be "No, it's not possible... Then someone showed me curl (7.22) doing it... Well, at least it throws a descriptive message when a website with a "bad" intermediate cert is hit. Has anyone gotten this test to work or should I continue to work on a patch for this to submit? Inquiring minds want to know Bruce Ferrell

2 2

Issue with alerts and DISPLAYGROUP
by Patrick.Dubois＠bac-lac.gc.ca 29 Aug '12

29 Aug '12

Hi, I'm using version 4.3.10 of Xymon and I'm experiencing issues with the DISPLAYGROUP parameter in alerts.cfg. The following is working well: DISPLAYGROUP=Sharepoint COLOR=yellow TIME=W:0700:1700 MAIL sharepointadmins at hotmail.com<mailto:sharepointadmins at hotmail.com> REPEAT=365d but the following is not working (not generating alerts): DISPLAYGROUP=Sharepoint COLOR=yellow EXSERVICE=cpu TIME=W:0700:1700 MAIL sharepointadmins at hotmail.com<mailto:sharepointadmins at hotmail.com> REPEAT=365d If I try to specify a service with SERVICE instead of EXSERVICE, I get the same behavior. SERVICE and EXSERVICE work as expected in the following scenario (but of course I'm missing the group filtering): HOST=* COLOR=yellow EXSERVICE=cpu TIME=W:0700:1700 MAIL sharepointadmins at hotmail.com<mailto:sharepointadmins at hotmail.com> REPEAT=365d I think it's a bug, unless I missed something in the documentation. Thanks! Patrick Dubois Administrateur de systèmes (serveurs) | System Administrator (servers) Services de gestion de l'infrastructure | Infrastructure Management Services Services partagés Canada @ Bibliothèque Archives Canada | Shared Services Canada @ Library and Archives Canada 550, boul. de la Cité, Gatineau QC K1A 0N4 | 550 de la Cité Blvd., Gatineau QC K1A 0N4 patrick.dubois at lac-bac.gc.ca<mailto:patrick.dubois at lac-bac.gc.ca> Téléphone | Telephone 613 899-9568 Télécopieur | Fax 819 994-6835 Gouvernement du Canada | Government of Canada [Description: Description: Description: http://publiservice.gc.ca/ssc-spc/images/dt-tg/ds-sm/ef-fe/SPC_signature_co…]

1 0

OpenVPN Monitoring
by r.oliver＠web.de 29 Aug '12

29 Aug '12

Dear All, I have tried to find a way to Monitor my OpenVPN Servers. The hard thing was that OpenVPN is a service that is running via UDP by default. So here is the script I've come up with: ----------- #!/bin/sh BBHTAG=openvpn # What we put in bb-hosts to trigger this test COLUMN=$BBHTAG # Name of the column, often same as tag in bb-hosts OPENVPNPORT=1194 $BBHOME/bin/bbhostgrep $BBHTAG | while read L do set $L # To get one line of output from bbhostgrep HOSTIP="$1" MACHINEDOTS="$2" MACHINE=`echo $2 | $SED -e's/\./,/g'` COLOR=green MSG="$BBHTAG status for host $MACHINEDOTS" #... do the test, perhaps modify COLOR and MSG PROGCHECK=`which nc` if [ -z "$PROGCHECK" ]; then MSG="${MSG} `echo "Program \"nc\" not found."` " fi /bin/echo -ne "\x38\x01\x01\x01\x01\x01\x01\x01\x01\x00\x00\x00\x00\x00" | /bin/nc $MACHINEDOTS $OPENVPNPORT -u -w 1 | /usr/bin/hexdump -v -e '"\\""x" 1/1 "%02x" ""' | $GREP -q '\\x01\\x01\\x01\\x01\\x01\\x01\\x01\\x01\\x00\\x00\\x00\\x00' if [ "$?" != "0" ]; then COLOR=red MSG="${MSG} `echo "OpenVPN offline"` " else MSG="${MSG} `echo "OpenVPN online"` " fi $BB $BBDISP "status $MACHINE.$COLUMN $COLOR `date` ${MSG} " done exit 0 ----------- Could someone try this out whether it works? Thanks in advance!

1 2