July 2010 - Xymon - Xymon mailman web

how to see graphs for last week/month for perticular test ?snapshot report
by pankaj.dorlikar＠gmail.com 11 Jun '12

11 Jun '12

hi, i have 2 problems.. 1)i want to see last month cpu utilization graph for the perticular host,how can i see that? 2)when i want to see snapshot report for say date 2dec09 7:30am,front pages(bb-html etc) shows time stamp of 2dec09 7:0am,but when i click to see status of perticular test of perticular host,it shows me status of 29oct09 12:40pm.how to overcome it? -- Pankaj V. Dorlikar

5 9

AIX df command problem
by sholmes42＠mac.com 13 Sep '11

13 Sep '11

I am having a problem similar to one recently posted, but this is on an AIX host and is caused (I think) by a problem in the way xymon does df on AIX. The xymon version is 4.2.3 and the AIX OS version is 3.5 (I think, I'm a Solaris Admin). Anyway, I can find no way to force the df on AIX to use POSIX formatting AND to report only local file systems. Even if I could, I don't think I can override what xymon uses for df since the comments in the hobbitclient.cfg file say that the envariable definition is only there for compatibility and it is not used. The problem is that an NFS mount file system line in the df output is being parsed wrong and the number in the usage column is being used for the percent used (Available) so that is a huge number. Not only that, again similar to the previous problem, the file system name is being picked up as the contents of the last 3 fields in the output line, i.e. it is not /mnt/tsmupdates (which is what it should be) but "3020485 50% /tsm/updates", which, of course, prevents me from effectively ignoring that file system. I've tried using DISK %.*/mnt/.* IGNORE and that doesn't work. Does anyone have any suggestions? We have a big red page where we shouldn't have. Thanks, Steve -- The test of a democracy is not the magnificence of buildings or the speed of automobiles or the efficiency of air transportation, but rather the care given to the welfare of all the people. -Helen Adams Keller, lecturer and author (1880-1968) Truth never damages a cause that is just. -Mohandas Karamchand Gandhi (1869-1948)

3 4

Alternate view for Management
by hballinger＠heritage-healthcare.com 31 Jan '11

31 Jan '11

Our support staff likes having our monitoring system setup & organized by logical category/type. However, management is more interested in a view that is organized by business impact. For example: IT Top Level View- Hosted Services Corporate Connectivity Corporate Servers Corporate Desktops Remote Laptops Facility Connectivity Facility Desktops Management Top Level View- Email (Subpage containing all dependent hosts/services) HRMS (same as above) Sales CRM i.e. a listing for each "service" provided - possibly even maintained below subpages for each business unit Management wants availability and impact by the specific service and business unit impacted (10,000 foot view showing all dependencies), while IT is interested in the detail of the cogs/gears that make it all work. Do I just create a second listing for all of the hosts on a separate page structure? Would this generate twice as many network tests for the hosts that are listed duplicate times? Any suggestions? Thanks!

7 10

bbcombotest Links
by me2unix＠gmail.com 22 Oct '10

22 Oct '10

Hi, Could not find anyone answering this request yet.. Here's a fix to add links to the bbcomtest reports page : Just change line 455 of hobbitd/bbcombotest.c >From : sprintf(msgline, "&%s %s\n", colorname(vwalk->color), vwalk->symbol); To: sprintf(msgline, "<a href=\"/hobbit-cgi/bb-hostsvc.sh?HOST=%s&SERVICE=%s\">&%s</a> %s\n", gethname(vwalk->symbol), gettname(vwalk->symbol) , colorname(vwalk->color), vwalk->symbol); recompile and have links on the color dots of the tested items within combotests. --Hezki ** - *To*: <hobbit (at) hswn.dk> - *Subject*: bbcombotest Links - *From*: "Aiello, Steve $GE, Corporate, consultant$" <steve.aiello(at) ge.com> - *Date*: Wed, 28 Feb 2007 10:43:21 -0500 - *Thread-index*: AcdbTzrE4X3/UsIMTSiBvQJbJOnunQ== - *Thread-topic*: bbcombotest Links ------------------------------ Hi all, I thought I saw this question asked before, but did see it in the mailling list. When using bbcombotests, is it possible to have the combotest report with hyperlinks to the actual tests ? If I see one of the statuses are red or yellow, my nature instict is to wonder why, and be able to drill down for details. thanks, Steve

5 4

Memory report faulty
by everett.vernon＠gmail.com 11 Oct '10

11 Oct '10

Hi all Another contract, another emial address. :-) Hope you all had a good festive season. Quick question. Was there ever a resolution to this issue? http://www.hswn.dk/hobbiton/2009/03/msg00036.html Searched the archive, and couldn't find one. To recap, this was where memory reports something bizzare, like this > Memory Used Total Percentage > red Physical 4294961768M 8192M 4294967229% > green Swap 1600M 8191M 19% Regards Vernon

4 6

Problem with disk monitoring
by shailesh.paudyal＠gmail.com 07 Oct '10

07 Oct '10

All, Please see below, there is a problem with disk monitoring on one of the server. Can some one tell me if I did something wrong? W]d Jul 28 10:34:31 EDT 2010 - Filesystems NOT ok 7% / (8816628% used) has reached the PANIC level (95%) 38% /u01 (90371708% used) has reached the PANIC level (95%) 2% /tmp (9254468% used) has reached the PANIC level (95%) 34% /usr (6261556% used) has reached the PANIC level (95%) 18% /opt (7775588% used) has reached the PANIC level (95%) 4% /var (13653064% used) has reached the PANIC level (95%) 3% /home (27514896% used) has reached the PANIC level (95%) 30% /boot (67864% used) has reached the PANIC level (95%) 14% /u02 (1697518148% used) has reached the PANIC level (95%) 94% /u03 (136865636% used) has reached the PANIC level (95%) Filesystem 10 4-b]ocks Used Available Capacity Mounted on /dev/sda9 9920592 591896 8816628 7% / /dev/sda10 152435112 54195172 90371708 38% /u01 /dev/sda8 9920592 154056 9254468 2% /tmp /dev/sda7 9920592 3146968 6261556 34% /usr /dev/sda6 9920592 1632936 7775588 18% /opt /dev/sda5 14877060 456092 13653064 4% /var /dev/sda3 29753588 702880 27514896 3% /home /dev/sda1 101086 28003 67864 30% /boot /dev/mapper/VolGroup02-u02 2064204960 261831260 1697518148 14% /u02 /dev/mapper/VolGroup03-u03 2064204960 1822483772 136865636 94% /u03 -- Shailesh K. Paudyal shailesh.paudyal at gmail.com

5 13

bb-services: regexes in expect string?
by AaronZink＠eharmony.com 21 Sep '10

21 Sep '10

I have this defined in bb-services: [smtp-transaction] send "helo hobbit.domain.com\r\nmail from: <test at domain.com>\r\nrcpt to: <test at anotherdomain.com>\r\nquit\r\n" expect "220-mx01.domain.com ESMTP\r\n220 ESMTP READY\r\n250 mx01.domain.com\r\n250 sender <test at domain.com> ok\r\n" port 25 and a host that uses it in bb-hosts. This works great, except that it's tied to the host mx01.domain.com in the expect string. Is it possible to use regexes in the expect string, such as expect "220-mx0[0|1].domain.com"? So far my attempts have been unsuccessful and the man pages don't seem to indicate if this is possible. Aaron Zink Manager, Corporate IT eHarmony.com 626.795.4814

3 2

ackcode in log file
by vadud3＠gmail.com 23 Aug '10

23 Aug '10

Does anyone know if any of the hobbit server log files contain the ackcode in it? I do not think it does by default. I could not find it in my limited query. I know SCRIPT has the ACKCODE. But I am asking if it is in any of the log file by default. If the answer is no, then what would it take to add the ackcode in notification log file ? I am guessing some code change on lib/notifylog.c 's do_notifylog function would be necessary along with some change in lib/notifylog.h. Please advise -- Asif Iqbal PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing?

3 5

Premature end of script headers: bb-histlog.sh
by vadud3＠gmail.com 19 Aug '10

19 Aug '10

I am getting core dump only for historical yellow data for some hosts. apache log: [Fri Jul 23 16:17:22 2010] [error] [client 192.168.0.94] Premature end of script headers: bb-histlog.sh, referer: https://xymon.example.net/hobbit-cgi/bb-hist.sh?HISTFILE=cer1cc-015.clientc… (hobbit)@xymon:~$ REQUEST_METHOD=GET SCRIPT_NAME=bb-histlog.sh QUERY_STRING="HOST=cer1cc-014.clientcare.example.net&SERVICE=cpu&TIMEBUF=Wed_Jul_21_14:56:42_2010" cgi-bin/bb-histlog.sh generates core dump. (hobbit)@xymon:~$ file core core: ELF 32-bit LSB core file 80386 Version 1, from 'hobbitsvc.cgi' (hobbit)@xymon:~$ gdb server/bin/hobbitsvc.cgi core [..] (gdb) backtrace #0 0xfed64fa1 in realfree () from /lib/libc.so.1 #1 0xfed655bf in cleanfree () from /lib/libc.so.1 #2 0xfed64adb in _malloc_unlocked () from /lib/libc.so.1 #3 0xfed64a04 in malloc () from /lib/libc.so.1 #4 0x08067621 in xmalloc (size=12) at memory.c:119 #5 0x0805a1f1 in expand_env (s=0x80707bb "4.2.3") at environ.c:296 #6 0x0805a08c in xgetenv (name=0x80715c9 "HOBBITDREL") at environ.c:165 #7 0x0805db08 in headfoot (output=0x8089e28, template=0x80717d3 "histlog", pagepath=0x80716c6 "", head_or_foot=0x80717f3 "header", bgcolor=4) at headfoot.c:1114 #8 0x0805f639 in generate_html_log (hostname=0x80913a8 "cer1cc-014.clientcare.example.net", displayname=0x80913a8 "cer1cc-014.clientcare.example.net", service=0x808aa90 "cpu", ip=0x8099168 "10.240.91.208", color=4, sender=0x80ad668 "10.240.91.208", flags=0x806ff21 "", logtime=0, timesincechange=0x8047bbc "0.00 minutes", firstline=0x8094128 "yellow Wed Jul 21 14:56:35 GMT 2010 up: 98 days, 0 users, 624 procs, load=7.51", restofmsg=0x80ae61f "&yellow Load is HIGH\nSystem clock is 0 seconds off\n\n\n\n\n 2:56pm up 98 days, 17:04, 0 users, load average: 9.38, 7.51, 4.34\n624 processes: 623 sleeping, 1 running, 0 zombie, 0 stopped\nCPU states: 0"..., acktime=0, ackmsg=0x0, acklist=0x0, disabletime=0, dismsg=0x0, is_history=1, wantserviceid=0, htmlfmt=0, hobbitd=0, multigraphs=0x8073060 ",disk,inode,qtree,quotas,snapshot,TblSpace,if_load,", linktoclient=0x808ae40 "bb-histlog.sh?CLIENT=cer1cc-014.clientcare.example.net&amp8TIMEBUF=1279724202", nkprio=0x0, nkttgroup=0x0, nkttextra=0x0, output=0x8089e28) at htmllog.c:150 #9 0x08054189 in do_request () at hobbitsvc.c:421 #10 0x0805530e in main (argc=4, argv=Cannot access memory at address 0x555202bc ) at hobbitsvc.c:518 Any one have a workaround? I tried to change the CFLAGS and recompile with gcc 4.3.4 replaced CFLAGS = -g -O2 -Wall -Wno-unused -D_REENTRANT $(LFSDEF) $(OSDEF) with CFLAGS = -g -O2 -Wall -D_REENTRANT $(LFSDEF) $(OSDEF) that did not help here. -- Asif Iqbal PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing?

2 4

Feature? svcs
by everett.vernon＠gmail.com 03 Aug '10

03 Aug '10

Hi all I have noticed, if you run BBWin, you get the svcs column, showing all the services running on the client. And, the server understands this test. I can add entries into hobbit-client.cfg along the lines of HOST=foo SVC "SomeService" startup=automatic status=started COLOR=red and if the service isn't running it will change to a red status. I want the same functionality for Solaris services. I wrote a very simple test for Solaris services (see below), but of course, that's a client-side test, and the services to alert on can't be configured at server side. (It will also only register services in maintenance mode, not disabled services) I thought maybe I could hack the hobbitclient-sunos.sh to send the stuff under the [svcs] heading, like BBWin does, but that didn't work. (I assume there is some logic defining what the server expects from each client type) I even went so far as to massage the output of svcs -aH to look at least similar to the BBWin svcs output. I would like to be able to have a Solaris client, sending service info, and be able to define entries in hobbit-client.cfg similar to this HOST=foo SVC /system/cron STATE=online COLOR=red And have the server bleat if that service isn't online. What would that require to make work? Regards Vernon #!/bin/ksh TEST="svcs" SVCS=/usr/bin/svcs TEMPFILE=$BBTMP/svcs.tmp COLOUR="green" date > $TEMPFILE echo >> $TEMPFILE $SVCS -x > $TEMPFILE.tmp EXIT=$? if [ $EXIT -eq 0 ] then # Command ran OK if [ -s $TEMPFILE.tmp ] then # It's not empty - bad COLOUR=red echo "There are issues with configured services." >> $TEMPFILE echo "Please investigate" >> $TEMPFILE echo >> $TEMPFILE else # It is empty - All is good. echo "All services running OK" >> $TEMPFILE echo >> $TEMPFILE fi else # Error running command COLOUR="yellow" echo "Error running $SVCS -x " >> $TEMPFILE echo "Please investigate" >> $TEMPFILE echo >> $TEMPFILE fi cat $TEMPFILE.tmp >> $TEMPFILE rm $TEMPFILE.tmp $SVCS -a >> $TEMPFILE $BB $BBDISP "status $MACHINE.$TEST $COLOUR `cat $TEMPFILE`" rm $TEMPFILE

3 9