November 2012 - Xymon - Xymon mailman web

Duplicate email Recipients in Info page
by Don.Kuhlman＠schawk.com 10 May '13

10 May '13

Does that mean it will stay green for a half hour before going red if the service has stopped ? Thanks! On 6/6/12 12:35 PM, "Betsy Schwartz" <betsy.schwartz at gmail.com> wrote: >On Wed, Jun 6, 2012 at 1:28 PM, Don Kuhlman <Don.Kuhlman at schawk.com> >wrote: > >> I've been working on the email issue and now seem to have another >>issue. I >> shut down the bbwin service on two windows computers that xymon was >> monitoring and it's not alerting that the service and/or process has >> stopped. > >Default purple interval , aka "status lifetime" is half an hour... I >think it's hardcoded on the server side. > >Individual alerts can also set their lifetimes. We do that with some >of our custom tests, run them every half hour with a lifetime of a >couple of hours. >

5 7

xymon e-mail alerts show up as attachments instead of e-mail
by icepickjazz＠gmail.com 29 Mar '13

29 Mar '13

Not a wild thought. Yes you are correct. I was getting ready to post my findings. We recently moved our xymon server from fedora core 3 to SLES 11. Fedora used mail and SLES 11 mail is linked to mailx. The windows clients are sending carriage returns in the status ^M (carriage return). The problem occured on SuSe server but not on redhat server, because mailx expects input text to be in Unix format, with lines separated by *newline*(^J, \n) characters only. Non-Unix text files that use *carriage return* (^M, \r) characters in addition will be treated as binary data. Hence the reason we get notifications that show up as attachments. Also explains why AIX servers don't have the problem and the windows servers do. Thanks On Mon, Sep 19, 2011 at 9:36 PM, Scott Pfister <icepickjazz at gmail.com>wrote: > Not a wild thought. Yes you are correct. I was getting ready to post my > findings. We recently moved our xymon server from fedora core 3 to SLES > 11. Fedora used mail and SLES 11 mail is linked to mailx. The windows > clients are sending carriage returns in the status ^M (carriage return). > The problem occured on SuSe server but not on redhat server, because mailx > expects input text to be in Unix format, with lines separated by *newline*(^J, \n) characters only. Non-Unix text files that use > *carriage return* (^M, \r) characters in addition will be treated as > binary data. Hence the reason we get notifications that show up as > attachments. Also explains why AIX servers don't have the problem and the > windows servers do. > > Thanks > > > On Mon, Sep 19, 2011 at 8:08 PM, Ralph Mitchell <ralphmitchell at gmail.com>wrote: > >> Wild thought - could the Windows reports be turned into attachments in >> your email server and/or client because the line-ends are <cr><lf> instead >> of just <lf> ?? >> >> Ralph Mitchell >> >> >> On Mon, Sep 19, 2011 at 6:06 PM, Scott Pfister <icepickjazz at gmail.com>wrote: >> >>> Good day, >>> >>> We are running xymon 4.2.3 and are experiencing some particularities with >>> e-mail alerts showing up as attachments like att00001.bin instead of the >>> normal message. The problem is sporadic. For example windows server throws >>> CPU alter. E-mail sent shows up as an attachment in the recipient e-mail. >>> AIX servers throw cpu alerts don't experience this problem. Client >>> shouldn't matter. Any tips for troubleshooting this issue? If we open the >>> attachment in the editor it shows what should be in the body of the email. >>> Hobbit server is running postfix. >>> >>> red Mon Sep 19 15:07:50 EDT 2011 [indyctx8.companyname.com] up: 10:12, 1 >>> users, 224 procs, load=100%, PhysicalMem: 3072MB(74%) >>> >>> >>> >>> Memory Statistics >>> Total Physical memory: 3220598784 bytes (3.00GB) >>> Available Physical memory: 850558976 bytes (812.00MB) >>> Total PageFile size: 4294967295 bytes (4.00GB) >>> Available PageFile size: 4294967295 bytes (4.00GB) >>> Total Virtual memory size: 2147352576 bytes (2.00GB) >>> Available Virtual memory size: 2106589184 bytes (1.96GB) >>> >>> Most active processes >>> 96.23% oexplore (0x26a4 [9892]) >>> 00.60% svchost (0xbf0 [3056]) >>> 00.32% lsass (0x23c [572]) >>> 00.26% oexplore (0x29c8 [10696]) >>> 00.25% oexplore (0x255c [9564]) >>> 00.21% Smc (0x3c4 [964]) >>> 00.20% SmcGui (0x3230 [12848]) >>> 00.19% oexplore (0x20dc [8412]) >>> 00.16% oexplore (0x14d0 [5328]) >>> 00.13% XTE (0xfc4 [4036]) >>> >>> See >>> http://hobbit.companyname.docm/hobbit-cgi/bb-hostsvc.sh?HOST=indyctx8&SERVI… >>> >>> >>> thanks, >>> >>> Scott >>> >>> >>> >>> _______________________________________________ >>> Xymon mailing list >>> Xymon at xymon.com >>> http://lists.xymon.com/mailman/listinfo/xymon >>> >>> >> >

3 2

ISO documentation on flapping status?
by betsy.schwartz＠gmail.com 05 Feb '13

05 Feb '13

We've got a bunch of servers showing as "flapping". I understand that this is a new feature but haven't found the docs on how to tweak it. Where and how can this be changed? thanks Betsy

3 3

holiday as planned downtime
by Jurgen.Claessens＠cegeka.be 11 Dec '12

11 Dec '12

Hey all, I'm running xymon 4.3.7 and I have been trying to find a way so that the holidays are being showed as what would be the same for the alerting. For me this would mean that if it's a holiday they would show as planned downtime and not show up in the allnongreen window. The alerting itself works without problems, but I cannot find a way for the planned downtime act the same way. Can anyone point me in the right direction? This e-mail and all files transmitted as attachment(s) thereto are confidential and solely intended for the individual to whom or the organization to which they are addressed. If you received this e-mail by mistake, please notify Cegeka's Service Desk at cegeka.support at cegeka.be or call +32 (0)11 240 363. We thank you in advance. Cegeka hereby confirms that this message has been swept by Sophos for the presence of viruses.

2 3

Solaris 10 xymon SMF
by Rolands＠logicaltech.com.au 30 Nov '12

30 Nov '12

Hi, There is a minor issue with the Solaris 10 SMF for xymon by Galen Johnson that I improved a bit. Thanks Galen for the nice work :D At least it works fine for us. When you disable the xymon service it could wait for up to 5 min before disabling it because the iostat and vmstat are running. (300 min) I been trying some different approaches in the manifest but it doesn't really work properly. I had to change the default timeout (30) to 300 to cater for the 300min possible run time of iostat, vmstat. Otherwise it will go into maintenance mode. Today I had a different approach and changed the xymon.sh script instead, that worked really nice. In server/bin/xymon.sh I added these simple lines to the stop function. (works in Solaris 10) "stop") if test -s /usr/local/xymon/logs/xymonlaunch.pid then kill -TERM `cat /usr/local/xymon/logs/xymonlaunch.pid` echo "Xymon stopped" else echo "Xymon is not running" fi rm -f /usr/local/xymon/logs/xymonlaunch.pid # kill off the remaining xymon processes, ie vmstat iostat etc. sleep 2 pkill 'iostat|vmstat' -U xymon ;; Maybe it could help someone else out there. I'm running 4.3.10 - Roland

3 3

FreeBSD 8.0+ memory reporting patch
by feld＠feld.me 30 Nov '12

30 Nov '12

An embedded and charset-unspecified text was scrubbed... Name: not available URL: <http://lists.xymon.com/pipermail/xymon/attachments/20121129/5d584af0/attach…>

3 2

MAPÉ - Monitoring App Page Extension
by cami＠hack.co.za 29 Nov '12

29 Nov '12

Howdy All, We're working on a new (animated) front-end for Xymon which allows you to have different views into your environment. It's called "MAPÉ" (pronounced: MAP-EE). The project was started because the default "Main View" / "Non-green" view is quite outdated and such a great monitoring system needs a new face. We would love to get some feedback on the current screenshots and the roadmap (preferably on the mailing list). Instead of subjecting everyone to screenshots, you can take a look at it: http://mape.hack.co.za The first milestone has been reached with a dynamic / animated tree and will be up for download once we got a bit of discussion going around where we should be headed. Regards, Cami

8 18

mysql password file on centos 5, xymon client
by leon＠one.co.il 29 Nov '12

29 Nov '12

Hi, I have a question. Recently I've added a script to monitor the mysql replication. I got it off the internet: ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- #!/bin/bash ############################################################################################ test -e /usr/local/mysql/bin/mysql && PROG=/usr/local/mysql/bin/mysql test -e /usr/bin/mysql && PROG=/usr/bin/mysql MAXDRIFT=5 ############################################################################################ MYSQLCOLOR="" MYSQLLINE="" HOSTIP=127.0.0.1 IPADDR=127.0.0.1 MACHINE=`uname -n` MAXSEC=60 MYSQLUSER=root MYSQLPORT=3306 LINE=`echo "show slave status \G" | ${PROG} -h ${HOSTIP} -u ${MYSQLUSER} -P ${MYSQLPORT} 2>&1` MASTERPOS=`echo $LINE | grep Read_Master_Log_Pos | awk -F 'Read_Master_Log_Pos: ' '{print $2}' | awk '{print $1}'` SLAVEPOS=`echo $LINE | grep -i Exec_Master_Log_Pos | awk -F 'Exec_Master_Log_Pos: ' '{print $2}' | awk '{print $1}'` SLAVEUP=`echo $LINE | grep Running | wc -l` SLAVE_SECOND_BEHIND_MASTER=`echo $LINE | grep 'Seconds_Behind_Master:' | awk -F 'Seconds_Behind_Master: ' '{print $2}' | awk '{print $1}'` test $SLAVEUP -eq "1" && DRIFT=$((${MASTERPOS} - ${SLAVEPOS})) echo "${LINE}" | ${GREP} "Running: No" >/dev/null if test "$?" -eq "0" then MYSQLCOLOR="red" MYSQLLINE="`${DATE}` MYSQL Slave in error ${LINE}" elif test ${SLAVEUP} -ne "1" then MYSQLCOLOR="green" MYSQLLINE="`${DATE}` MYSQL Slave is Stopped ${LINE}" elif test ${SLAVE_SECOND_BEHIND_MASTER} -ge ${MAXSEC} then MYSQLCOLOR="yellow" MYSQLLINE="`${DATE}` MYSQL Slave is ${DRIFT} statements behind! ${LINE}" else MYSQLCOLOR="green" MYSQLLINE="`${DATE}` MYSQL Slave is ${DRIFT} statements behind! ${LINE}" fi #echo "${BB} ${BBDISP} status ${MACHINE}.mysqlslave ${MYSQLCOLOR} ${MYSQLLINE}" ${BB} ${BBDISP} "status ${MACHINE}.mysqlslave ${MYSQLCOLOR} ${MYSQLLINE}" ----------------- To make it working, you need to create the file : ~hobbit/.my.cnf [mysql] password=XXXXXXXX [mysqlclient] password= XXXXXXXX [mysqladmin] password= XXXXXXXX [mysqlshow] password= XXXXXXXX [mysqldump] password= XXXXXXXX ------------------------------------------- chown hobbit: ~hobbit/.my.cnf chmod 600 ~hobbit/.my.cnf Add the section in clientlaunch.cfg [mysqlslave] ENVFILE $HOBBITCLIENTHOME/etc/hobbitclient.cfg CMD $HOBBITCLIENTHOME/ext/mysqlslave.sh LOGFILE $HOBBITCLIENTHOME/logs/mysqlslave.log INTERVAL 5m Restart hobbit-client. ------------------------------------------------------------------------------------------------------------------------ as you can see, the author instructs to put the .my.cnf file with all the passwords in the ~hobbit/.my.cnf directory. While that went fine on the Ubuntu 12.04 server, it doesn't work on the CentOS 5 server. I got the xymon client.rpm from some website and it just doesn't recognize this file no matter where I put it. The version is 4.3.10. the client is installed in these directories: /etc/xymon-client /usr/libexec/xymon-client /usr/share/xymon-client I also noticed that there's no /var/lib/xymon-client directory. The Ubuntu version was installed via apt-get and still has "hobbit" naming. On CentOS client I renamed all the "hobbit" into "xymon". If I put the pass into the script itself - it works. But I really don't want to do that. Any ideas? Thanks, Lenny

3 4

[Newbie] Smaller, lighter Xymon based program?
by kfc.android＠gmail.com 28 Nov '12

28 Nov '12

Hi guys, I'm very new to this Xymon-land, spent some time to install and configure Xymon on my virtual machine runs CentOS 6.3 Here's my thought, since I have very limited skill in this monitoring field, my target is only to have a program which can test the 'ping' only. Is that doable if I can use Xymon and disable some of the functions? Has anyone ever think about this before? Maybe some hint or guides, thanks! Btw, I'm working on this as my school project, hopefully I didn't break/offend anyone's work and copyright, if so, please let me know. Many thanks :D -- Best, James

7 10

Combining Macros in alerts.cfg EXHOST?
by betsy.schwartz＠gmail.com 28 Nov '12

28 Nov '12

I'm trying to make my alerts.cfg a lot simpler. I've got two macros defining groups of hosts. To simplify slightly: $NOPAGE=%(^*.eng|^netmon2).example.com $VIP=%(^web|^ftp|^mail-relay).example.com,web-vip,mail-vip Each macro works by itself when I use it in a rule. However I want to make a rule that groups these two together, something like so: HOST=* EXHOST=%($VIP,$NOPAGE) SERVICE=$SYSENG include syseng.escalation.cfg I'm not getting this to work , with various combinations : %($NOPAGE,$VIP) %($NOPAGE|$VIP) The rules are catching hosts in $NOPAGE but not in $VIP I could make a third combined group, but since I've got other rules that just apply to $VIP or $NOPAGE it would be nice to be able to do this thanks for any clues Betsy

3 4