Apache RRD 🔗 link

Hi user-6b5f82a4758d@xymon.invalid,

a new version is out and it's running fine...

But my Apache problem exists further.

I can not monitoring the Apache-Status.

If I into the configuration the following line insert:

xxx.xxx.xxx.xxx    www.xxxxxxxxx.com # ssh
http://www.xxxxxxx.com/xxxxx.htm
apache=http://www.xxxxxxx.com/server-status?auto
LARRD:*,apache:apache|apache1|apache2|apache3

then hobbit say's -> "BBTEST "red" Programm crashed Fatal signal
caught!" But the LARRD graph for Apache its running but all other tests
not!

Whats Wrong ??? 

I use Hobbit 4.0.3 on SuSE Linux 9.2

The gdb say's (bbtest-net) 

#4  0x401a6d41 in raise () from /lib/tls/libc.so.6
#5  0x401a8529 in abort () from /lib/tls/libc.so.6
#6  0x0805ca12 in sigsegv_handler (signum=11) at sig.c:57
#7  <signal handler called>

#7544 0x4000d6ae in _dl_name_match_p () from /lib/ld-linux.so.2
Previous frame inner to this frame (corrupt stack?)

(bbgen)

#4  0x401a6d41 in raise () from /lib/tls/libc.so.6
#5  0x401a8529 in abort () from /lib/tls/libc.so.6
#6  0x0805ca12 in load_all_links () at links.c:130
#7  <signal handler called>

#7544 0x4000d6ae in _dl_name_match_p () from /lib/ld-linux.so.2
Previous frame inner to this frame (corrupt stack?)

Please help.....


Cheers...

Alexander Sprotte
Alexander Sprotte [mailto:user-76113fbca463@xymon.invalid]

On Mon, May 23, 2005 at 01:39:23PM +0200, user-76113fbca463@xymon.invalid wrote:

Yes, there are a couple of bugs that haven't been dealt with in
4.0.3 - because I decided that they were so specific to your setup
that I wouldn't hold up the release any more.

then hobbit say's -> "BBTEST "red" Programm crashed Fatal signal
caught!"

Try adding "--no-ares" to the bbtest-net commandline in
hobbitlaunch.cfg. Does that help ?

If not, I'd like to see the output from

  bbcmd bbtest-net --debug --ping --checkresponse

because the gdb trace is so messed up I cannot figure out 
where the real problem happens.


Regards,
Henrik

Hi....

No, there is the part of the hobbitlaunch.cfg

[bbnet]
        ENVFILE /home/hobbit/server/etc/hobbitserver.cfg
        NEEDS hobbitd
        CMD bbtest-net --no-ares --report --ping --checkresponse
        LOGFILE $BBSERVERLOGS/bb-network.log
        INTERVAL 5m

If not, I'd like to see the output from

Here is it:

  bbcmd bbtest-net --debug --ping --checkresponse
2005-05-23 14:03:12 Using default environment file
/home/hobbit/server/etc/hobbitserver.cfg
Aborted (core dumped)

I use "gdb bin/bbtest-net tmp/core" is it right ??


Regards,
Alex

On Mon, May 23, 2005 at 02:11:25PM +0200, user-76113fbca463@xymon.invalid wrote:

Hrm, something funky going on with the option passing there.
Try this instead:

* login as the hobbit user
* run "bbcmd" (this starts a sub-shell with all of the Hobbit
  environment defined).
* run "bbtest-net --debug --ping --checkresponse --no-ares"


Henrik

On Mon, May 23, 2005 at 02:11:25PM +0200, user-76113fbca463@xymon.invalid wrote:

Perfectly correct, but there is something happening inside bbtest-net that messes up the call stack, so it is no longer
possible to see what happened.


Henrik

The test is running and crash.

2005-05-23 14:20:15 Adding to combo msg: status www1,xxxxxx,com.http
green Mon May 23 14:20:05 2005: OK
2005-05-23 14:20:15 Recipient listed as '10.129.11.101'
2005-05-23 14:20:15 Standard BB protocol on port 1984
2005-05-23 14:20:15 Will connect to address 10.129.11.101 port 1984
2005-05-23 14:20:15 Connect status is 0
2005-05-23 14:20:15 Sent 32981 bytes
2005-05-23 14:20:15 Closing connection
Aborted (core dumped)

Alex

Hi Henrik,

I have now add the Apache Test on the local Apache where Hobbit is running and its running fine. No crash on problem... :)

But the remote test crashed :(

Alex

On Mon, May 23, 2005 at 01:39:23PM +0200, user-76113fbca463@xymon.invalid wrote:

xxx.xxx.xxx.xxx    www.xxxxxxxxx.com # apache=http://www.xxxxxxx.com/server-status?auto

then hobbit say's -> "BBTEST "red" Programm crashed Fatal signal caught!"

With some help from Alexander I got this sorted out - it appears
to be a bug that triggers when the response to the server-status
request is very large (> 32 KB data).

-- 
Henrik Storner