November 2015 - Xymon - Xymon mailman web

Monitoring process with regex - fails with 4.3.4 and BBWin 0.13
by sirusirus14＠gmail.com 28 Nov '15

28 Nov '15

Hello I will use the BBWin 0.13 in Windows Server 2012, the Xymon 4.3.4 on CentOS I use. In order to monitor the regular expression to the Windows process, we set as follows in the BBWin.cfg of Windows Server. <procs> <setting name "notepad" rule = "= 2" alarmcolor = "red" /> <setting name "% notepad C: \ Users \ Administrator \ Desktop \ test.txt" rule = "= 2" alarmcolor = "red" /> <setting name "% notepad *" rule = "= 2" alarmcolor = "red" /> <setting name "% ^ notepad *" rule = "= 2" alarmcolor = "red" /> </ procs> The first line "notepad" only I've correctly match, the rest regular expression does not match. Please tell me if there is a mistake. Thank you very much.

2 1

4.3.21 AIX client -> 4.3.10 Linux server ?
by abs＠shadymint.com 27 Nov '15

27 Nov '15

Hi, We have built a new(ish) xymon package for AIX 7.1 and the client works fine sending updates via xymonclient-aix.sh. However, using xymonnet we see no results for network tests. The xymonnet binary works without error and --debug shows it sending the expected results. Wireshark confirms this is received on the server but as a type 'extcombo'. This is a long shot :- Changes from 4.3.12 -> 4.3.13 (08 Jan 2014) =========================================== * rev 7339 * New "extcombo" message type allows for combo messages of all types (usually "status" and "data"). So, being lazy rather than looking through the code at this time, can anyone confirm if the client should be reverting to combo when the server cannot accept extcombo and is there any way to prevent the client from using extcombo untill we can get the server upgraded? Will peruse the code this weekend if I get chance, just hoping someone can shortcut the investigation for me ... Thanks -- Andy

2 2

Critical Systems editor -- which characters are problematic in text input fields?
by hobbit＠elyograg.org 25 Nov '15

25 Nov '15

I'm running Xymon 4.3.23 with the http status patch from 4.3.24. I've been adding some info to the Critical Systems config. I have learned that the double quote character (") causes text in fields like Instructions to be truncated when the entry is brought up for further editing. The full text was properly saved in critical.cfg, including the double quote and everything after it. There hasn't been a new alarm since setting it up, so I don't know whether the full message would have appeared on the test status page when accessed from Critical Systems. Are there any characters that will cause actual problems? I think it's probably a good idea to pop up an error when submitting the form if illegal characters are present, or maybe just have a note on the page that indicates which characters should not be used. If having a double quote in the message is not actually problematic, then I believe there's a bug. Thanks, Shawn

1 0

Xymongrep suggestion: pick up additional test parameters
by glauber.ribeiro＠experian.com 25 Nov '15

25 Nov '15

There are xymon tests which include additional parameters. For example: the network tests can include a port number, and apache has a url. Currently, xymongrep doesn't support these, because it only works with exact matches. It would be nice for it to be able to recognize colon (tag:parameter or tag:parameter1,parameter2) and equals (tag=parameter or tag=parameter1,parameter2) and supply these as additional words in the output. Just an idea, in case it's considered useful for a future version. glauber

3 3

Xymon 4.3.24 Released
by cleaver＠terabithia.org 24 Nov '15

24 Nov '15

Hello all, Xymon 4.3.24 has been released and is available on SF. This is a bug fix release that corrects the processing of HTTP Status codes in xymonnet. Outside of a few specific common HTTP codes, most of the rest were being flagged incorrectly as Critical errors, instead of Warnings or OK. The test now works as originally intended. All users are encouraged to upgrade. Regards, Japheth "J.C." Cleaver Xymon v4 Maintainer

2 4

"Discarding timed-out partial msg" Error Messages
by matt1299＠gmail.com 24 Nov '15

24 Nov '15

Hey all, Lately, I've been seeing quite a few error messages show up in xymond indicating that it was discarding a timed-out partial message from some machine. i.e. Latest error messages: Discarding timed-out partial msg from X.X.X.X They seem to be happening sporadically but more often than usual as of late. Maybe one or two every couple of days or so. They don't seem to be coming from the same machine/machines either. Is this something I should be worried about? Are there any side-effects from this happening too much? What are the causes of this happening? Any way to make it not happen as much? Any ideas or advice is greatly appreciated! Thanks!! -- Matt Vander Werf

2 1

Graphing terminal server connection issue
by rectifier＠gmail.com 24 Nov '15

24 Nov '15

Hi all, I am having an issue with graphing a powershell script that announces how many connections we have to RDS farm environment. The html output (which contains the info I want to graph) is as such: </style> </head><body> <h2>Terminal Server Status</h2><br><b>Total users: 23<br>Active Sessions: 19 <br>Disconnected Sessions: 4</b><br><br><table class="NormalTable"><colgroup><col /><col /><col /><col /><col /><col /></colgroup><tr><th>Username</th><th>Session Name</th><th>Session ID</th><th>State</th><th>Idle Time (hh:mm)</th><th>Login etc....................... This is all on same line and there is a lot more info about the table and CSS. Is CSS ok with rrd graphing and Xymon? In xymonserver.cfg, I have termusers=ncv and termusers listed in GRAPHS= At the end of the file (just to simplify as it wasn't working) I have NCV_termusers="ActiveSessions:GAUGE,*:NONE" In graphs.cfg (but there is no rrd file as yet) I have: [termusers] TITLE Terminal Server Users YAXIS Users DEF:active=termusers.rrd:ActiveSessions:AVERAGE DEF:disc=termusers.rrd:DisconnectedSessions:AVERAGE LINE2:active#00CCCC:Active Users LINE2:disc#FF0000:Disconnected Users COMMENT:\n GPRINT:active:LAST:Active Users \: %5.1lf%s (cur) GPRINT:active:MAX: \: %5.1lf%s (max) GPRINT:active:MIN: \: %5.1lf%s (min) GPRINT:active:AVERAGE: \: %5.1lf%s (avg)\n GPRINT:disc:LAST:Disconnected Users \: %5.1lf%s (cur) GPRINT:disc:MAX: \: %5.1lf%s (max) GPRINT:disc:MIN: \: %5.1lf%s (min) GPRINT:disc:AVERAGE: \: %5.1lf%s (avg)\n But it won't create the rrd file and I am pretty sure this is because of the output, there is something in there that creates an issue. 1) any tips on how to diagnose this type of problem more efficiently? 2) anyone had a similar issue when graphing from POWERSHELL or similar to mine? The old termusers scripts (possibly from Xymonton) used to graph properly but when my colleague made the PowerShell version, the graphing never worked. Thanks!

3 4

IBM Blade Server template
by jsaharan＠outlook.com 24 Nov '15

24 Nov '15

Does anyone have a working IBM blade server devmon template? Thanks & Regards Jasbir Saharan

1 0

http test -- 202 status red after upgrade from 4.3.21 to 4.3.23
by hobbit＠elyograg.org 23 Nov '15

23 Nov '15

I have an http test that returns a 202 response, which was perfectly acceptable when the xymon server was running 4.3.21. Today I upgraded it to 4.3.23, and now that 202 response is generating a red alarm. Here's a relevant excerpt from the 4.3.22 section in the release notes: ============ The default rules for HTTP response codes have been simplified and made more generic: 1xx = Warning (when the final code received) 2xx = OK 3xx = Warning (except 302/303/307 = OK) 4xx = Critical 5xx = Critical ============ If I switch to httpstatus and configure "2.." for acceptable status codes, the alarm goes away. I think what happens with the regular http test is either a bug in xymon or a documentation problem, and I'm curious what the devs think. Thanks, Shawn

3 5

alerts includes
by Galen.Johnson＠sas.com 23 Nov '15

23 Nov '15

Hey, I seem to be having a weird issue I'm hoping someone else can confirm. I'm us?ing "directory" to manage my alerts configs. For example directory /opt/xymon/server/etc/alerts.d HOST=* MAIL personal at example.com DURATION>10 REPEAT=30 RECOVERED However, it doesn't seem to actually get to the HOST entry. It uses the last file listed in the alerts.d folder as the catchall. I would never have noticed this had I not been having a test of some new servers using a script intended for another group. Can anyone else confirm this behavior? I'm currently using 4.3.21. thanks =G=

3 5