September 2009 - Xymon - Xymon mailman web

hobbitd_channel uses 99% cpu
by wendell＠enflight.com 29 Sep '09

29 Sep '09

Hi, I'm running xymon-4.3.0-beta1 on a CentOS 5.3 system. For quite a while xymon has been running smoothly. However, lately hobbitd_channel has been using 99% of the cpu. Sometimes there are two hobbitd_channels running, each using 49% of the cpu. Is there an ipc channel or shared memory that is somehow stuck or corrupted? ipcs shows nothing when xymon is stopped, and shows 9 memory segments an 9 semaphore arrays when xymon is running. What is going on? Did I mis-configure something? Is there a way to watch it or debug the problem? Thanks, Wendell

1 0

bbcmd fails
by william.wagner＠ipacc.com 29 Sep '09

29 Sep '09

How can I get bbcmd to work? I am doing this as hobbit user on a Solaris xymon 4.2.3 server: bbcmd <default port> --channel=<some client> cat the result: 2009-09-29 13:51:47 Using default environment file <path>/etc/hobbitserver.cfg 2009-09-29 13:51:47 execvp() failed: No such file or directory thanks

2 1

Vertical columns. Is it possible?
by camorae＠googlemail.com 29 Sep '09

29 Sep '09

Hi guys, does anybody know, if its possible to make the "check columns" at the vertical and not horizontal? Its kinda of unhandy when you allways have to scroll to the left when you want to see all your checks. Thanks, Cayo

9 23

Load threasholds when bbwin in 'central' mode
by Craig_Whilding＠mentor.com 29 Sep '09

29 Sep '09

I decided to try out 4.3 beta 2 along with the central mode for 'bbwin' clients. Unfortunately when doing this I discovered many of the windows boxes were alerting red because the load was higher than my defaults of 4 and 8... Now I've upped these to 95 and 100 to represent the percentage the windows boxes are using but this means the unix boxes could be broken and I wont know without setting individual alert thresholds. Is there a way to change the windows 'load' to be 'load percentage' or something so each type of box is using a different monitoring type? I know there were discussions about having UNIX boxes sending in a load percentage but that this was a difficult thing to achieve correctly. Our hostnames are normally dictated by service rather than platform unfortunately so I can't always just put so8/w2k etc pcre expressions in the thresholds to differentiate OS type. Regards, Craig Whilding

3 2

[hobbit] DNS Test - No Record Lookup
by Craig.Cook＠carquest.com 29 Sep '09

29 Sep '09

>##############The bb-hosts file entry is as follows: > >page solaris Solaris >192.168.X.X solaris # Name:"SOLARIS" dns nis ntp telnet pop3 pop3s imap imaps You are not using FQDN. Try (as the xymon user on the xymon host) dig @solaris solaris You may not be resolving correctly. Confirm /etc/resolv.conf is correct Try adding your domain to bb-hosts 192.168.X.X solaris.example.com # Name:"SOLARIS" dns nis ntp telnet pop3 pop3s imap imaps Also try removing ' Name:"SOLARIS" ' (I don't think that will matter though) Craig

7 16

Files to move to a new server?
by d.tom.schmitt＠L-3com.com 28 Sep '09

28 Sep '09

I will be moving the existing system from a desktop PC to a Dell 1850 Server over the next week or so. In order to preserve the current information, what files/directories do I need to move to the new system? What permissions do I need to worry about? I want to keep the RRD history data for the graphs also. I am running Xymon 4.3.0-0.beta2. Is there a newer version that I should be running? Thanks, Tom Schmitt Senior IT Staff - R&D Phone (801) 594-3030

3 4

Incomplete mdstat output
by josh＠imaginenetworksllc.com 28 Sep '09

28 Sep '09

Since Nov 30 2007 this server has had raid green excluding two red events - on on the 19th and then just a few minutes ago. They read identical to normal with the exception of the very top. Green: *green md0 Status OK green md2 Status OK green md1 Status OK* and the first red *red md1 Status OK* and the second red *red md2 Status OK* Now both the red and green show the informative part identically: ============================ /proc/mdstat =========================== Personalities : [raid6] [raid5] [raid4] [raid1] md0 : active raid1 sdb1[1] sda1[0] 104320 blocks [2/2] [UU] md2 : active raid5 sdc1[2] sdb2[1] sda2[0] 4192768 blocks level 5, 256k chunk, algorithm 2 [3/3] [UUU] md1 : active raid5 sdc2[2] sdb3[1] sda3[0] 151894016 blocks level 5, 256k chunk, algorithm 2 [3/3] [UUU] unused devices: ============================ End of file ============================ In the past I have found these as legit red/yellow causes: Yellow... green md0 Status OK green md1 Status OK green md2 Status OK yellow md3 resync in progress Red... green md0 Status OK green md1 Status OK green md2 Status OK red md3 Disk Failed Any idea as to why mdstat doesn't say the top part isn't saying anything? This may be more of a ##linux question but I was hoping someone had seen this before. Josh Luthman Office: 937-552-2340 Direct: 937-552-2343 1100 Wayne St Suite 1337 Troy, OH 45373 "When you have eliminated the impossible, that which remains, however improbable, must be the truth." --- Sir Arthur Conan Doyle

1 0

Disk growth in MB during graphs timespan, anybody done it?
by Stephen.Mullin＠usa.net 28 Sep '09

28 Sep '09

Interested in showing delta/growth of filesystems, in MB, not % as is default. Someone must have done this by now... ok if it's just in the legend and not a line in graph. This way on my monthly reports you get both visual as shown currently, but for the time span shown, the delta in MB is shown.

1 0

HOBBITMENU issue on xymon 4.3.0 debian stable
by pkc_mls＠yahoo.fr 28 Sep '09

28 Sep '09

Hi all, I'm running xymon client and server on debian stable (5.0). Both were installed via the packages. (ie no compilation). I have an error in the log that comes each 5 minutes in the bb-display.log: xgetenv: Cannot find value for variable HOBBITMENU The menu is replaced with &HOBBITMENU. I searched through the source code, but I couldn't find any reference to this HOBBITMENU. Has anyone already solved this ? thanks.

2 2

tests: files,ports,procs
by camis＠mweb.co.za 28 Sep '09

28 Sep '09

Hi all, I'm in the process of migrating from the bbclient to the xymon/hobbit client and everything is pretty straight forward so far. However when i'm using the client i'm seeing some tests that are showing as white "files, ports, procs" which isn't a bad thing but i'd like to turn it off completely. Is the only way of doing it with using: NOCOLUMNS:files,ports,procs ? I have hundreds of hosts to monitor and having to specify that for each and every host seems very clunky so i'm guessing i missed something obvious that allows me to disable it globally for all hosts. Any ideas? Regards, Cami News24 Headlines - SAFA war not yet over? (2009-09-28 08:19) view on http://bit.ly/GZce1 - Shot from behind in farm attack (2009-09-28 08:05) view on http://bit.ly/2lS6fH - World Cup's cash cow (2009-09-28 08:17) view on http://bit.ly/awtrx www.news24.com

2 2