- Xymon - Xymon mailman web

hobbit and unbound dns
by gabrielescu＠gmail.com 24 Apr '09

24 Apr '09

i have to monitorize the unbound dns activity. there 2 plugins working with Cacti and Munin ( http://www.unbound.net/documentation/howto_statistics.html) which are not helpping me too much. do you tried to use hobbit and Unbound dns do you have some scripts? Any help is appreciated:) Gabi

1 0

Disk Status graphs
by isolderj＠hotmail.com 24 Apr '09

24 Apr '09

I am running xymon 4.2.3. My graphs for the disk are showing all the partitions in the trends but in the disk status graph does not contain all the partitions. I have searched on the forums but have found no fix for this behaviour. is there anything that I can do to fix this so the Disk Status Graph looks the same as the trends graph. Thanks _________________________________________________________________ Rediscover Hotmail®: Now available on your iPhone or BlackBerry http://windowslive.com/RediscoverHotmail?ocid=TXT_TAGLM_WL_HM_Rediscover_Mo…

2 6

Include graph only in trends column
by nelis＠nlr.nl 24 Apr '09

24 Apr '09

Hello, For a few servers, we would like to add one graph to the trends column only, a graph showing the total disk space used at that server. An additional graph definition is created to build the sum of the used space on (a subset of) the disks: [disktua] TITLE Total disk usage YAXIS Disk space used [B] DEF:d0=disk,shared,asd,pool0001.rrd:used:AVERAGE DEF:d1=disk,shared,asd,pool0002.rrd:used:AVERAGE DEF:d2=disk,shared,asd,pool0003.rrd:used:AVERAGE DEF:d3=disk,shared,asd,pool0004.rrd:used:AVERAGE DEF:d4=disk,shared,asd,pool0005.rrd:used:AVERAGE DEF:d5=disk,shared,asd,pool0006.rrd:used:AVERAGE CDEF:tdu=d0,d1,+,d2,+,d3,+,d4,+,d5,+,1024,* LINE1:tdu#00CC00:Total GPRINT:tdu:MIN:Min\: %5.1lf %sB GPRINT:tdu:MAX:Max\: %5.1lf %sB GPRINT:tdu:AVERAGE:Avg\: %5.1lf %sB GPRINT:tdu:LAST:Cur\: %5.1lf %sB\n If the configuration for this server in bb-hosts is modified to contain TRENDS:*,disk:disk|disk1|disktua the graph is shown. However, as there are three graphs to show both "disk" and "disk1", there are one graph and two links visible for "disktua". This result in understandable, but it does not look nice. If the configuration for this server is modified to contain TRENDS:*,disk:disk|disk1,disktua graph "disktua" is not shown at all. Is there a way to include a graph in the trends column only? Kind regards, Wim Nelis. ******************************************************************************************************* The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages. *******************************************************************************************************

4 3

Xymon 4.2.3 Compile Issues HP-UX 11.31 CLOCK_MONOTONIC
by John.Browning＠sas.com 23 Apr '09

23 Apr '09

Hi, I'm compiling a 4.2.3 client binary for HP-UX 11.31 and am receiving the following error: mefunc-client.o timefunc.c timefunc.c: In function 'gettimer': timefunc.c:55: error: 'CLOCK_MONOTONIC' undeclared (first use in this function) timefunc.c:55: error: (Each undeclared identifier is reported only once timefunc.c:55: error: for each function it appears in.) timefunc.c: In function 'getntimer': timefunc.c:67: error: 'CLOCK_MONOTONIC' undeclared (first use in this function) gmake[1]: *** [timefunc-client.o] Error 1 gmake[1]: Leaving directory `/root/xymon-4.2.3/lib' gmake: *** [lib-client] Error 2 After some research I see numerous threads and posts on a slew of sights stating CLOCK_MONOTONIC is not supported in HP-UX nor MAC OSX, both of which I will need to compile binaries for. Any help or tips would be greatly appreciated. Thanks! -------------------------------------------- John Browning UNIX Systems Administrator R&D DataCenter www.sas.com<http://www.sas.com/> SAS(r) ... THE POWER TO KNOW(r)

2 3

DEPENDS tag strangeness
by SKreien＠la-z-boy.com 23 Apr '09

23 Apr '09

Hi everyone, Has anyone ever figured out the problem where using a depends tag will cycle the test that is being depended on red-clear-red-clear when it fails? I saw posts to the list in the archive but never found if there was a solution. Also, is there any way to do that same functionality with other non-network tests? Thanks, Scot Kreienkamp La-Z-Boy Inc. skreien at la-z-boy.com 734-242-1444 ext 6379

1 0

Still around - sort of - but a 4.3.0 beta 2 will be out tomorrow
by henrik＠hswn.dk 23 Apr '09

23 Apr '09

Hi everyone. I haven't quite dropped off the planet, but some work got seriously in the way of Xymon developments. As some of you have noticed a "4.4.0" tree appeared on the Sourceforge subversion archive - some of the features planned for 4.4.0 was needed immediately, and I didn't want that to mess up the 4.3.0 code, so I have been trying to balance both sets of releases at one time. The short version of that is "don't use the 4.4.0 tree yet, unless you can tolerate lock-ups and crashes". I'll be wrapping up the next beta version of 4.3.0 later tonight or early tomorrow and put a tar-file on Sourceforge. This has the bug- fixes which have been reported on the mailing list (and fixes quite a few bugs I found by myself). The documentation is still not quite up-to-date, but that is all that is missing. I then expect to be completely off-line until May 12, so you have two weeks to test it. And if anyone feel like submitting bugfixes or improving on the docs, you are very welcome. Regards, Henrik PS: It's a real joy to see how well the community support works on this mailing list. A big "thanks!" to all of you helping out.

2 1

Remote worker module?
by scochran＠univaud.com 23 Apr '09

23 Apr '09

(Sorry if these questions are already covered somewhere. I'm a newbie, and all of my searching didn't turn anything up.) I'm working on an integration with Hobbit for a client. They are using version 4.2.0. I need to be able to process incoming data on a server that is remote to the Hobbit server. I read in the hobbitd_channel man page that you can locate worker modules remotely by using the --net argument. Something like: bin/hobbitd_channel --channel=data --log=$BBSERVERLOGS/remote_data.log --net 10.10.4.67:12345 should (I believe) send "data" messages to host 10.10.4.67 on port 12345. However when I try this, I just get error messages in my log saying: Our child has failed and will not talk to us: Channel data, PID 3251, cause Bad file descriptor (My remote client is, for now, just a netcat listener on the port. I can telnet to it, so I know that it's there and listening.) So my questions are: 1) Does this remote capability exist in version 4.2? 2) Are there any special requirements of the client such that a generic listener (like netcat) would provoke this error message? 3) Is there a sample (or real) worker module I could look at, that is capable of being remoted in this way? Or what's the remote equivalent of the get_hobbitd_message() function call? any insight or examples would be appreciated. thanks, steve --------------------------------------------------------------------- Notice from Univa UD Postmaster: This email message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. This message has been content scanned by the Univa UD Tumbleweed MailGate. ---------------------------------------------------------------------

2 2

[hobbit] clients with different host name than uanme -n for xymon-4.2.2
by bgmilne＠staff.telkomsa.net 22 Apr '09

22 Apr '09

----- "pankaj dorlikar" <pankaj.dorlikar at gmail.com> wrote: > hello, > i am using different hostname for clients than uname -n.hence i run > runcklient.sh with --hostname=myhostname and it is running fine and > sending data to server.But now whenever in future if runclient.sh gets > stopped or client process gets killed,is ther any way to avoid running > runclient.sh with -hostname=myhostname mannually?where sholud i fix so > that my runclient.sh script always get"myhostname" and not the uname > -n.i dont want to change bb-hosts file CLIENT=hostname tag. > Any suggestions? If you are using an init script, such as the one supplied in debian/hobbit-client.init or rpm/hobbit-client.init, then you would be able to set: CLIENTHOSTNAME=myhostname in /etc/default/hobbit-client (or, /etc/sysconfig/hobbit-client in more Red Hat-like RPMs). Regards, Buchan

1 0

clients with different host name than uanme -n for xymon-4.2.2
by pankaj.dorlikar＠gmail.com 22 Apr '09

22 Apr '09

hello, i am using different hostname for clients than uname -n.hence i run runcklient.sh with --hostname=myhostname and it is running fine and sending data to server.But now whenever in future if runclient.sh gets stopped or client process gets killed,is ther any way to avoid running runclient.sh with -hostname=myhostname mannually?where sholud i fix so that my runclient.sh script always get"myhostname" and not the uname -n.i dont want to change bb-hosts file CLIENT=hostname tag. Any suggestions? -- Pankaj V. Dorlikar

1 0

Sys log (UNCLASSIFIED)
by mathew.t.rogers＠us.army.mil 22 Apr '09

22 Apr '09

Classification: UNCLASSIFIED Caveats: NONE I recently implemented Hobbit/xymon 4.2, and have a question. I looked through the archives, and didn't find what I was looking for, so if this question has been answered previously, I apologize. My supervisor wants all server logs to stay on the web page for approximately 90 days. The desire is to use Hobbit as a syslog server as well as a monitoring tool. It seems to me that when new information is sent to the Hobbit server, all the old information is removed from the web page and put in the data directory. The question is, Is there a way to make the logs visible on the webpage (when you click on msgs for each specific server) for 90 days, while still updating the logs hourly? Thanks in advance, Mathew Rogers Classification: UNCLASSIFIED Caveats: NONE

3 2