August 2011 - Xymon - Xymon mailman web

httpstatus inconsistency (yellow masks red)
by troy＠athabascau.ca 08 Aug '11

08 Aug '11

I just found the time to test this on Xymon 4.3.4 and the behavior is the same as 4.3.2; a yellow from analysis.cfg masks the red state. Any help would be appreciated. cheers, Troy -- This communication is intended for the use of the recipient to whom it is addressed, and may contain confidential, personal, and or privileged information. Please contact us immediately if you are not the intended recipient of this communication, and do not copy, distribute, or take action relying on it. Any communications received in error, or subsequent reply, should be deleted or destroyed. ---

1 0

another issue after upgrading .. custom temperate monitor extension not recording proper history..
by tsztur＠gmail.com 08 Aug '11

08 Aug '11

I have a temperature sensor extension that does not get recorded in the "100 events received in the past 201 minutes" section any longer after running a test to change the status to red. The status icon turns RED and the page is sent however after the status goes back to normal (GREEN) the "Last 50 log entries" still goes GREEN for the last 4 days. This is the same page that turned RED. Also , the Current non-green Systems history shows nothing of the entry turning RED at all. All other systems seem to be recording except for these custom ones. any idea why? This is a temperature sensor and has a custom script (in the ./ext ).. however the sensor works fine however the history is not being recorded properly. any ideas?

1 0

upgraded from 4.3.0-beta to 4.3.4 .. [info] tab no longer works "internal server error"
by tsztur＠gmail.com 08 Aug '11

08 Aug '11

error_log says "Premature end of script headers: svcstatus.sh," the permissions seem fine. All other pages that call up svcstatus.sh seem to work except the Info tab. Any idea why?

3 2

Spikes in activity.
by henrik＠hswn.dk 08 Aug '11

08 Aug '11

On 21-07-2011 15:51, Neil Simmonds wrote: > I’ve looked through the documentation for Xymon and I can’t seem to find > any way to only have an icon colour change if a threshold has been > passed for x number of consecutive tests. > > I’m aware that alerts allow this kind of thing but would really like the > dashboard not to change colour if we just get a spike in disk, cpu or > memory usage (amongst other things). > > Is there any way to do this? It's been possible for network tests for many years, but that was really a kludge on top of the old Big Brother system. So your mail provided an excellent excuse for getting this implemented in a generic fashion (so I can deprecate the old network-test-specific method). The attached patch implements two new settings you can put into hosts.cfg: "delayred" and "delayyellow". So to have the "cpu" status not go red until after 15 minutes and disk status not go red until after half an hour, you would use this: 10.0.0.1 host1.example.com # delayred=cpu:15,disk:30 Patch applies against 4.3.4 - older versions will need some manual tweaking to make it into the xymond.c file. Or wait for 4.3.5. Regards, Henrik

3 5

Looking for "page too old" script
by josh＠imaginenetworksllc.com 08 Aug '11

08 Aug '11

Got it working! Thanks!!! FILE /var/www/inxwireless/network/output.png red mtime<1800 This is correct. From what the documentation states, of which I may be understanding it incorrectly, this is backwards. Initially I did mtime>1800 based on what I read on the man page. The test went red "File was modified 31 seconds ago - should be >1800" Josh Luthman Office: 937-552-2340 Direct: 937-552-2343 1100 Wayne St Suite 1337 Troy, OH 45373 On Fri, Jul 29, 2011 at 4:31 PM, Tim McCloskey <tm at freedom.com> wrote: > > Do I put this under [linux]? If I do this, will it only have the clients > with an appropriate FILE check or ask this file of every linux box? > > Just took another look at a working config....here's a couple notes. > You don't need the whole [linux] class, you can use a [hostname] in > client-local.cfg > ... > [your-hostname] > file:/var/www/foo/bar/output.png > ... > > The stanza above with a matching FILE directive from hobbit-clients.cfg > (both server side) works for me, on 4.2.0. > > If you have this setup - sounds like you do - and it's not working you can > Try touching the logfetch.$foo.cfg and logfetch.$foo.status files in the tmp > dir on the client (create them however you need, just make sure to chown > them to your hobbit user). > > You should not need to populate the files, just create them. Example > file... > cat someclienthostname:~hobbit/client/tmp/logfetch.someclienthostname.cfg > log:/some/path/to/somefile:10240 > file:/bin/su > > Tim > > >

3 3

Double combo's - returning -1 - xymon 4.3.3
by david.gore＠verizon.com 07 Aug '11

07 Aug '11

Henrik, In regards to double combobs, since no one is replying I suppose either no one is using this feature or it is broken? As a side note: xymoncmd xymond_alert --test host service --duration=29 --color=red 00002332 2011-08-07 12:57:58 *** Match with 'EXHOST=$HOSTS_EXCLUDED HOST=* SERVICE=service COLOR=red' *** 00002332 2011-08-07 12:57:58 Matching host:service:page 'host:service:Site/Platform,Place' against rule line 432 00002332 2011-08-07 12:57:58 Failed 'SCRIPT=$SENDPAGE $PAGE_NMS DURATION>30m REPEAT=365d' (min. duration 1740<1800) The duration above is in minutes when clearly the documentation specifies seconds? --test HOST SERVICE [options] Shows which alert rules matches the given HOST/SERVICE combination. Useful to debug configu- ration problems, and see what rules are used for an alert. The possible options are: --color=COLORNAME The COLORNAME parameter is the color of the alert: red, yellow or purple. --duration=SECONDS The SECONDS parameter is the duration of the alert in seconds. --group=GROUPNAME The GROUPNAME paramater is a groupid string from the analysis.cfg file. ~David -----Original Message----- From: xymon-bounces at xymon.com [mailto:xymon-bounces at xymon.com] On Behalf Of David W Gore Sent: Friday, July 29, 2011 19:17 To: xymon at xymon.com Subject: Re: [Xymon] Double combo's - returning -1 - xymon 4.3.3 Does anyone out there using double leveled combos, combos of combos that work? If so perhaps you could share an example from your config that works? Thanks in advance, Here is the example from the man pages: ---- WebCluster.http = (Web1.http || Web2.http) AppSrvCluster.procs = (AppSrv1.conn && AppSrv1.procs) || (AppSrv2.conn && AppSrv2.procs) Customer.cluster = WebCluster.http && AppSrvCluster.procs The first line defines a new test, with hostname "WebCluster" and the columnname "http". It will be green if the http test on either the "Web1" or the "Web2" server is green. The second line defines a "procs" test for the "AppSrvCluster" host. Each of the AppSrv1 and AppSrv2 hosts is checked for "conn" (ping) and their "procs" test. On each host, both of these must be green, but the combined test is green if that condition is fulfilled on just one of the hosts. The third line uses the two first tests to build a "double combined" test, defining a test that shows the overall health of the system. ---- So in theory it should work? For reference we are using Xymon 4.3.3 on RedHat Linux Enterprise 5.7. ~David From: xymon-bounces at xymon.com [mailto:xymon-bounces at xymon.com] On Behalf Of Rager, Richard R (Rich) Sent: Thursday, June 30, 2011 19:02 To: xymon at xymon.com Subject: [Xymon] Double combo's - returning -1 - xymon 4.3.3 Hi, Anyone using double combo messages? I was following the man pages, which state they are possible, but my tests calculate "-1" and the color comes out $unknown. Thu Jun 30 18:43:24 2011 (iam.db&&gam.db) = (-1&&-1) = 1 &unknown iam.db &unknown gam.db Sample config from combo.cfg. I have changed hostnames to protect the innocent...combos based on the values returned by dbcheck.pl...our DB's are served on 4 node Oracle Grid clusters, which is why we have alias' for the DB, such as servername_dbname. Hostnames are defined in hosts.d as "0.0.0.0 gam # noconn nonongreen noproppurple:*" gam/iam/impact all defined the same in hosts. iam.db = (servername\_iam.audit && servername\_iam.conn && servername\_iam.extent && servername\_iam.info && servername\_iam.invobj && servername\_iam.locks && servername\_iam.rollback && servername\_iam.session && servername\_iam.tblspace && servername\_iam.trends && servername\_iam.uptime) gam.db = (servername\_gam.audit && servername\_gam.conn && servername\_gam.extent && servername\_gam.info && servername\_gam.invobj && servername\_gam.locks && servername\_gam.rollback && servername\_gam.session && servername\_gam.tblspace && servername\_gam.trends && servername\_gam.uptime) These two above work fine. They are single layer combo tests...All my single layer combo tests work. The double combo is this: impact.db = (iam.db && gam.db) - same with our without parenthesys...if I use single ampersand, color propogates up, but staus still shows &unknown. We have tried multiple combinations taking the combo down to barest form, and still get the errors (&unknown). Log doesn't really help... combostatus.log.1:21316 2011-06-30 03:58:03 Symbolic 'event.db&&gam.db&&iam.db&&jbpm.db' converted to '-1&&-1&&-1&&-1' I put the --debug on the cmd in tasks, and not logging much else helpful. Appreciate some thoughs or suggestions. Rich Rager VZB ________________________________________ This message was created by the sender using your @mci.com address. If you wish to continue to receive messages from this sender, please respond informing them to use your @verizonbusiness.com email address. The mci.com email domain will be decommissioned October 31, 2010. Please inform sender to use your @verizonbusiness.com address, if you desire to continue to receive messages from them. Thank You Verizon Business Messaging _______________________________________________ Xymon mailing list Xymon at xymon.com http://lists.xymon.com/mailman/listinfo/xymon

1 0

use of aberrant behavior detection
by Mario.Contreras＠tigo.com.co 06 Aug '11

06 Aug '11

Thank you Henrik for your answer. But that file is not present in my installation: / # ls /opt/xymon/server/etc bb-hosts Groups hobbitgraph.cfg bb-hosts.20101224 hobbit-alerts.cfg hobbitGroups bb-services hobbit-apache.conf hobbitlaunch.cfg bbcombotest.cfg hobbit-clients.cfg hobbitpasswd client-local.cfg hobbit-nkview.cfg hobbitserver.cfg columndoc.csv hobbit-nkview.cfg.bak old_bb-hosts cookies.session hobbitcgi.cfg Mario Contreras -----Mensaje original----- De: xymon-bounces at xymon.com [mailto:xymon-bounces at xymon.com] En nombre de Henrik Størner Enviado el: domingo, 31 de julio de 2011 04:39 Para: xymon at xymon.com Asunto: Re: [Xymon] use of aberrant behavior detection > We are using xymon/devmon pair and we want to use for some tests the > "aberrant behavior detection" of RRDTOOL. Could someone please tell me > what changes are necessary in XYMON configuration files to do so? I > saw a similar question in the mail list back in 2005 but with no answer. You need to change the "rrddefinitions.cfg" file to include the RRDtool definitions for the Holt-Winters datasets. Since the rrd-definitions are set when the rrd-file is created, you must delete any existing rrd-files before this change takes effect. Regards, Henrik _______________________________________________ Xymon mailing list Xymon at xymon.com http://lists.xymon.com/mailman/listinfo/xymon Este mensaje (incluyendo cualquier anexo) contiene información confidencial que se encuentra protegida por la ley. La información que contiene sólo puede ser utilizada por la persona o la compañía a la cual está dirigido. Si usted no es un receptor autorizado, o por error recibe el mensaje, omita su contenido y elimínelo inmediatamente. Cualquier retención, difusión, distribución y/o copia así como la ejecución de cualquier acción basado en el contenido del mismo, se encuentra estrictamente prohibido, y esta amparado por la ley. -------------- This message (including any attachments) contains confidential information intented for a specific individual and purpose, and is protected by law. If you are not the intented recipient, you should delete this message. Any disclosure, copying, or distribution of this message, or the taking of any action based on it, is strictly prohibited.

2 1

Xymon version 4.3.4 available
by henrik＠hswn.dk 05 Aug '11

05 Aug '11

Hi, I have just uploaded version 4.3.4 of Xymon to Sourceforge, and it should be available from all of the mirror-sites shortly. This version fixes some rather nasty bugs, which could lead to data corruption of client messages (typically this would trigger wrong data for various statuses and spurious creation of disk RRD files) - I would like to thank Greg Shea for working with me to diagnose and fix this problem. Another important bugfix resolves the issue of acknowledgment cookies not being recognised after a while. Several other bugs that could cause various crashes, lockup and other nastiness were also fixed - see the full list below - so I strongly recommend that you upgrade to this version. Regards, Henrik Changes from 4.3.3 to 4.3.4: * Fix crashes and data corruption in Xymon worker modules (xymond_client, xymond_rrd etc) after handling large messages. * Fix xymond lock-up when renaming/deleting hosts * Fix xymond cookie lookup mechanism * Webpages: Add new HOSTPOPUP setting to control what values from hosts.cfg are displayed as a "comment" to the hostname (either in pop-up's or next to the hostname). * Fix xymond_client crash if analysis.cfg contains invalid configuration entries, e.g. expressions that do not compile. * Fix showgraph CGI crash when legends contain colon. * xymonnet: Include hostname when reporting erroneous test-spec * CGI utils: Multiple potential security fixes involving buffer- overruns when generating responses. * CGI utils: Fix crash when invoked with HTTP "HEAD" * CGI utils: Fix crashes on 64-bit platforms due to missing prototype of "basename()" function. * svcstatus CGI: Dont crash if history log is not a file. * Critical systems view CGI: Cross-site scripting fix * Fix recovery-messages for alerts sent to a GROUP * RRD "memory" status handler now recognizes the output from the bb-xsnmp.pl module (for Cisco routers). * Web templates modified so the menu CSS can override the default body CSS. * Acknowledge web page now allows selecting minutes/hours/days * Enable/Disable webpage enhanced, so when selecting multiple hosts the "Tests" column only lists the tests those hosts have.

3 8

Apache tag not working, not playing nice with http tag.
by betsy.schwartz＠gmail.com 04 Aug '11

04 Aug '11

I think I misunderstood how it was supposed to work. I was trying very very hard to get an apache status dot. So apache statuses show up in trends only? I confess I can't reproduce the nan values myself, after leaving it alone for a solid week. Is there perhaps a delay before those values start to show up? Maybe I was being too impatient. thanks Betsy On Sun, Jul 31, 2011 at 5:51 AM, Henrik Størner <henrik at hswn.dk> wrote: >> I've been trying to get apache results and am getting nothing, no >> results, no errors, as if it the tag wasn't there. >> >> UNLESS, and this is the weird part - *IF* I disable http for a given >> server I get *stats* under apache on the trends page, but still no >> apache column on the server page. >> >> Disabling http is not an option, as we've got nonstandard ports getting >> tested. >> >> With any http test enabled, just get blank "nan" values for apache. >> And no matter what, we don't get an apache column. > > The data for the Apache stats don't generate an Apache status - it just > generates an apache RRD file, you can see it in the "trends" column. > > I've tried reproducing your problem here, but I cannot - no matter how I try > to configure hosts.cfg, when running the network tests I get both the normal > "http" status update, and the "apache" data update. > > Could you try running the network tests against that host like this: > > xymoncmd xymonnet --noping --debug --no-update HOSTNAME 2>&1 >log.txt > > and see if there is a line with "data HOSTNAME.apache" together with the > apache statistics in there ? > > > Regards, > Henrik > _______________________________________________ > Xymon mailing list > Xymon at xymon.com > http://lists.xymon.com/mailman/listinfo/xymon >

1 0

Specifying port for BBDISP
by olivier＠audry.fr 03 Aug '11

03 Aug '11

hello how can I specify a different port for BBDISP in hobbitserver.cfg ? thx for your help oau

2 1