-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Dear all,
is it possible with Hobbit 4.2.0 to time alerts? I was asked this by colleagues. They want something that if a yellow alert is displayed for more than a configurable time that Hobbit/XYmon sets the state of this alert to red. E.g. if the msgs test is continuously yellow for 1 day or week the server automatically sets the state to red after this time. In the company where I operate the monitoring solution we have plenty of yellow alerts which get more or less ignored. So if an alert is not resolved within a certain time it should go red and tease the admins until they solve the problem.
If it is not possible with Hobbit 4.2.0 maybe it could be integrated in one of the next releases?
Second question: I have Hobbit server set up with centralised configuration. But some of the clients are running with local configuration. What do I have to do to monitor the log files like /var/log/messages on this client or check for the existence of certain files. At the moment I have the entries
"LOG /var/log/messages error" and "FILE /somedir/core NOEXIST" in localclient.cfg on the client and also on the server in hobbit-clients.cfg. But nothing is monitored.
Any help would be appreciated.
Cheers Torsten
+---------------------------------------------------------+ | Richter IT-Consulting | | E-mail : bb4 [at] richter-it.net | | | | Homepage: http://www.richter-it.net/ | +---------------------------------------------------------+ Download my public key from: http://www.keyserver.net:11371/pks/lookup?op=get&search=0x899093AC -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3-nr1 (Windows XP)
iD8DBQFJQD6i7DlmxomQk6wRAg9rAJ9Tv3XTvfZG0/rcM2B1y1A2tz7QgwCglbu4 nL52qBiklNhdVBbzISARewA= =UPDY -----END PGP SIGNATURE-----
In <49403EA2.7090906 at richter-it.net> Torsten Richter <bb4 at richter-it.net> writes:
is it possible with Hobbit 4.2.0 to time alerts? I was asked this by colleagues. They want something that if a yellow alert is displayed for more than a configurable time that Hobbit/XYmon sets the state of this alert to red.
Not possible today. It has not come up until now, but it's an interesting idea.
I guess some people would also like a status to go blue if it's been red for more than a day :-)
Second question: I have Hobbit server set up with centralised configuration. But some of the clients are running with local configuration. What do I have to do to monitor the log files like /var/log/messages on this client or check for the existence of certain files. At the moment I have the entries
"LOG /var/log/messages error" and "FILE /somedir/core NOEXIST" in localclient.cfg on the client and also on the server in hobbit-clients.cfg. But nothing is monitored.
Off the top of my head (meaning: I haven't tried it) I would think that you need to put something like log:/var/log/messages:10240 into ~hobbit/client/tmp/logfetch.cfg . The "10240" is the max amount of logdata to send to the Hobbit server (in kB).
Regards, Henrik
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hi Henrik,
thanks for the answers.
Henrik Størner wrote:
In <49403EA2.7090906 at richter-it.net> Torsten Richter <bb4 at richter-it.net> writes:
is it possible with Hobbit 4.2.0 to time alerts? I was asked this by colleagues. They want something that if a yellow alert is displayed for more than a configurable time that Hobbit/XYmon sets the state of this alert to red.
Not possible today. It has not come up until now, but it's an interesting idea.
It was just a thought since Hobbit/XYmon is recording everything with a timestamp and this timestamp "just" needs to be evaluated.
I guess some people would also like a status to go blue if it's been red for more than a day :-)
Please don't say that as it might be possible that my colleagues are reading this mailing list too and then they are going to tease me about integrating this kind of stuff :-/
Second question: I have Hobbit server set up with centralised configuration. But some of the clients are running with local configuration. What do I have to do to monitor the log files like /var/log/messages on this client or check for the existence of certain files. At the moment I have the entries
"LOG /var/log/messages error" and "FILE /somedir/core NOEXIST" in localclient.cfg on the client and also on the server in hobbit-clients.cfg. But nothing is monitored.
Off the top of my head (meaning: I haven't tried it) I would think that you need to put something like log:/var/log/messages:10240 into ~hobbit/client/tmp/logfetch.cfg . The "10240" is the max amount of logdata to send to the Hobbit server (in kB).
More or less it was right. I needed to create a logfetch.uname -n.cfg
in ~hobbit/client/tmp and put the following lines in it to make it work:
log:/var/log/messages/10240 ignore MARK file:/somedir/core
After that everything went fine.
Regards, Henrik
Thanks a lot for your help and the excellent work on this monitoring solution.
Regards Torsten
+---------------------------------------------------------+ | Richter IT-Consulting | | E-mail : bb4 [at] richter-it.net | | | | Homepage: http://www.richter-it.net/ | +---------------------------------------------------------+ Download my public key from: http://www.keyserver.net:11371/pks/lookup?op=get&search=0x899093AC -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3-nr1 (Windows XP)
iD8DBQFJQacW7DlmxomQk6wRAm3mAJ930XW2OPFDtWCJNIFNLDIYpVzodACeMzC9 ep7+i3mQ9Dbe7TuZnuVCnfo= =dhDB -----END PGP SIGNATURE-----
On Thu, Dec 11, 2008 at 5:49 PM, Torsten Richter <bb4 at richter-it.net> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hi Henrik,
thanks for the answers.
Henrik Størner wrote:
In <49403EA2.7090906 at richter-it.net> Torsten Richter <bb4 at richter-it.net> writes:
is it possible with Hobbit 4.2.0 to time alerts? I was asked this by colleagues. They want something that if a yellow alert is displayed for more than a configurable time that Hobbit/XYmon sets the state of this alert to red.
Not possible today. It has not come up until now, but it's an interesting idea.
It was just a thought since Hobbit/XYmon is recording everything with a timestamp and this timestamp "just" needs to be evaluated.
I guess some people would also like a status to go blue if it's been red for more than a day :-)
Please don't say that as it might be possible that my colleagues are reading this mailing list too and then they are going to tease me about integrating this kind of stuff :-/
Actually, it's not hard to make something go blue after a certain length of time - I have that set up on a machine that's pinging some F5 BigIP group members.
What you do is, get the list of red things using hobbitdboard, then pipe it into a loop that checks the date/time. If it's greater than your designated max, send a "disable until OK" message.
I'd post a copy of my script, except that the company laid me off at the beginning of last week, and I didn't think to keep a copy at home... If anyone's interested I can recreate it from scratch...
Ralph Mitchell
participants (3)
-
bb4＠richter-it.net
-
henrik＠hswn.dk
-
ralphmitchell＠gmail.com