Xymon Mailing List Archive search

Trend Graphs Stopped Showing Data

18 messages in this thread

list Mike Livenspargar · Mon, 08 Oct 2007 08:38:50 -0700 ·
We are a new installation of Hobbit 4.2.0, running on SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and Users
and Processes trend graphs for a particular machine stopped
showing data. All other trend graphs for that machine appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those two
graphs.
 
I'm not familiar with rrd but I do see that the procs.rrd,
users.rrd and la.rrd files on the hobbit server have
datetime stamps coinciding with the last reported data
(about 16 hours ago). The other .rrd files have current
datetime stamps. I've tried restarting the hobbit server and
the client but that did not fix the problem.
 
What steps can I take to try to resolve this problem?
 
Thank you,
Mike
list Tom L. Stewart · Mon, 8 Oct 2007 10:51:35 -0500 ·
I have the exact same problem on a Solaris system talking to Hobbit on
Solaris. My conclusion is that I need to reboot the client box to reset
the missing information on a who -b and uptime. Because the box is
production, it still won't get rebooted for another month. If you do try
the system reboot, please let me know either way on whether it worked or
not.
 
Tom
quoted from Mike Livenspargar


	From: Mike Livenspargar [mailto:user-c6d9d5761955@xymon.invalid] 
	Sent: Monday, October 08, 2007 10:39 AM
	To: user-ae9b8668bcde@xymon.invalid
	Subject: [hobbit] Trend Graphs Stopped Showing Data
	
	
	We are a new installation of Hobbit 4.2.0, running on SuSE Linux
ES 10 SP1 machines. All has been working fine for a couple of months
until yesterday when the CPU load and Users and Processes trend graphs
for a particular machine stopped showing data. All other trend graphs
for that machine appear to be working correctly. The graphs appear on
the Trends page, but the last 16 hours or so show no data for those two
graphs.
	 
	I'm not familiar with rrd but I do see that the procs.rrd,
users.rrd and la.rrd files on the hobbit server have datetime stamps
coinciding with the last reported data (about 16 hours ago). The other
.rrd files have current datetime stamps. I've tried restarting the
hobbit server and the client but that did not fix the problem.
	 
	What steps can I take to try to resolve this problem?
	 
	Thank you,
	Mike
list Henrik Størner · Mon, 8 Oct 2007 18:01:02 +0200 ·
quoted from Mike Livenspargar
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar wrote:
We are a new installation of Hobbit 4.2.0, running on SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and Users
and Processes trend graphs for a particular machine stopped
showing data. All other trend graphs for that machine appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module crashed
on a specific status message, which could also explain it. In that case,
you should have a core file in the ~hobbit/data/tmp/ directory,
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik
list Mike Livenspargar · Mon, 08 Oct 2007 09:44:26 -0700 ·
The CPU status for the host shows green. All other statuses
show green, and all the other trend graphs show current
data.
 
I don't see a core file, and nothing that looks like a log
entry in hobbitlaunch.cfg.
 
Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
quoted from Henrik Størner
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar
wrote:
We are a new installation of Hobbit 4.2.0, running on
SuSE
Linux ES 10 SP1 machines. All has been working fine for
a
couple of months until yesterday when the CPU load and
Users
and Processes trend graphs for a particular machine
stopped
showing data. All other trend graphs for that machine
appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those
two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being
updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module
crashed
on a specific status message, which could also explain it.
In that case,
you should have a core file in the ~hobbit/data/tmp/
directory,
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik
list Mike Livenspargar · Tue, 09 Oct 2007 12:45:27 -0700 ·
Well, the trend graphs are showing data again. The server
machine was rebooted but it appears the graphs were showing
data prior to the reboot. If I had to guess I'd think that
some process running on the client caused the data parsing
on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can
examine that might help me figure this out?
 
 I'll keep my eye on this to see what happens next.
quoted from Mike Livenspargar
 
Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar
wrote:
We are a new installation of Hobbit 4.2.0, running on
SuSE
Linux ES 10 SP1 machines. All has been working fine for
a
couple of months until yesterday when the CPU load and
Users
and Processes trend graphs for a particular machine
stopped
showing data. All other trend graphs for that machine
appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those
two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being
updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module
crashed
on a specific status message, which could also explain it.
In that case,
you should have a core file in the ~hobbit/data/tmp/
directory,
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik
list Josh Luthman · Tue, 9 Oct 2007 16:27:21 -0400 ·
Did you try restarting the hobbit server with the provided script in
~/server/hobbit.sh ?
quoted from Mike Livenspargar

On 10/9/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid> wrote:
 Well, the trend graphs are showing data again. The server machine was
rebooted but it appears the graphs were showing data prior to the reboot. If
I had to guess I'd think that some process running on the client caused the
data parsing on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can examine that
might help me figure this out?

 I'll keep my eye on this to see what happens next.

Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar wrote:
We are a new installation of Hobbit 4.2.0, running on SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and Users
and Processes trend graphs for a particular machine stopped
showing data. All other trend graphs for that machine appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module crashed
on a specific status message, which could also explain it. In that case,
you should have a core file in the ~hobbit/data/tmp/ directory,
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik

-- 

Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent it, poorly.
--- Henry Spencer
list Mike Livenspargar · Tue, 09 Oct 2007 13:39:18 -0700 ·
Yes, at least a couple of times before the reboot, to no
avail. And the boot startup scripts start the server with
the server/hobbit.sh script.
 
Mike 
user-4c45a83f15cb@xymon.invalid 10/9/2007 1:27:21 PM >>>
quoted from Josh Luthman
Did you try restarting the hobbit server with the provided
script in ~/server/hobbit.sh ?

On 10/9/07, Mike Livenspargar < user-c6d9d5761955@xymon.invalid>

wrote:Well, the trend graphs are showing data again. The
quoted from Josh Luthman
server machine was rebooted but it appears the graphs were
showing data prior to the reboot. If I had to guess I'd
think that some process running on the client caused the
data parsing on the server to go awry and once that process
ended things started working again. Is there any
hobbit-saved data I can examine that might help me figure
this out? 
 
 I'll keep my eye on this to see what happens next.
 
Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar
wrote:
We are a new installation of Hobbit 4.2.0, running on
SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and
Users
and Processes trend graphs for a particular machine
stopped
showing data. All other trend graphs for that machine
appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those
two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being
updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module
crashed
on a specific status message, which could also explain it.
In that case,
you should have a core file in the ~hobbit/data/tmp/
directory, 
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik


-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent
it, poorly. 
--- Henry Spencer
list Josh Luthman · Tue, 9 Oct 2007 16:57:08 -0400 ·
Just be certain, you are using the hobbit.sh script created upon
installation - correct?

Josh
quoted from Mike Livenspargar

On 10/9/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid> wrote:
 Yes, at least a couple of times before the reboot, to no avail. And the
boot startup scripts start the server with the server/hobbit.sh script.

Mike
user-4c45a83f15cb@xymon.invalid 10/9/2007 1:27:21 PM >>>
Did you try restarting the hobbit server with the provided script in
~/server/hobbit.sh ?

On 10/9/07, Mike Livenspargar < user-c6d9d5761955@xymon.invalid> wrote:
 Well, the trend graphs are showing data again. The server machine was
rebooted but it appears the graphs were showing data prior to the reboot. If
I had to guess I'd think that some process running on the client caused the
data parsing on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can examine that
might help me figure this out?

 I'll keep my eye on this to see what happens next.

Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar wrote:
We are a new installation of Hobbit 4.2.0, running on SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and Users
and Processes trend graphs for a particular machine stopped
showing data. All other trend graphs for that machine appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module crashed
on a specific status message, which could also explain it. In that case,
you should have a core file in the ~hobbit/data/tmp/ directory,
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik

--
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent it, poorly.
--- Henry Spencer
-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent it, poorly.
--- Henry Spencer
list Mike Livenspargar · Tue, 09 Oct 2007 15:27:54 -0700 ·
Yes, that is correct. The only way we start the server or
the clients is using the scripts created during
installation.
 
Mike
user-4c45a83f15cb@xymon.invalid 10/9/2007 1:57:08 PM >>>
quoted from Josh Luthman
Just be certain, you are using the hobbit.sh script created
upon installation - correct?

Josh

On 10/9/07, Mike Livenspargar < user-c6d9d5761955@xymon.invalid>

wrote:Yes, at least a couple of times before the reboot, to
quoted from Josh Luthman
no avail. And the boot startup scripts start the server with
the server/hobbit.sh script.
 
Mike 
user-4c45a83f15cb@xymon.invalid 10/9/2007 1:27:21 PM >>>
Did you try restarting the hobbit server with the provided
script in ~/server/hobbit.sh ?

On 10/9/07, Mike Livenspargar < user-c6d9d5761955@xymon.invalid
wrote: Well, the trend graphs are showing data again. The
server machine was rebooted but it appears the graphs were
showing data prior to the reboot. If I had to guess I'd
think that some process running on the client caused the
data parsing on the server to go awry and once that process
ended things started working again. Is there any
hobbit-saved data I can examine that might help me figure
this out? 
 
 I'll keep my eye on this to see what happens next.
 
Mike
user-ce4a2c883f75@xymon.invalid 10/8/2007 9:01:02 AM >>>
On Mon, Oct 08, 2007 at 08:38:50AM -0700, Mike Livenspargar
wrote:
We are a new installation of Hobbit 4.2.0, running on
SuSE
Linux ES 10 SP1 machines. All has been working fine for a
couple of months until yesterday when the CPU load and
Users
and Processes trend graphs for a particular machine
stopped
showing data. All other trend graphs for that machine
appear
to be working correctly. The graphs appear on the Trends
page, but the last 16 hours or so show no data for those
two
graphs.
Is the CPU status for that host showing a purple status ?
If yes, then for some reason the cpu status is not being
updated,
and this affects the graph updates.

There's also been one report that the hobbitd_rrd module
crashed
on a specific status message, which could also explain it.
In that case,
you should have a core file in the ~hobbit/data/tmp/
directory, 
and some log-entries in hobbitlaunch.cfg.


Regards,
Henrik


-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent
it, poorly. 
--- Henry Spencer 


-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent
it, poorly. 
--- Henry Spencer
list Mike Livenspargar · Wed, 10 Oct 2007 10:09:02 -0700 ·
And now, just as magically as the CPU Load and Users and
Processes graphs started showing data, they have stopped
again. All other trend graphs are functioning. As far as I
can tell all the data is coming over from the client and it
appears to be some issue with getting the data where it
needs to go to show up in the trend graphs, but I'm just
guessing here.
 
Are there any thoughts on what I could look at to try to
resolve this problem?
 
Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
quoted from Mike Livenspargar
Well, the trend graphs are showing data again. The server
machine was rebooted but it appears the graphs were showing
data prior to the reboot. If I had to guess I'd think that
some process running on the client caused the data parsing
on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can
examine that might help me figure this out?
 
 I'll keep my eye on this to see what happens next.
 
Mike
list Josh Luthman · Wed, 10 Oct 2007 13:51:29 -0400 ·
Have you tried checking to see if there are any quota or disk issues?  If
you were able to reboot the machine and get your RRD activity back up while
restarting the process did not fix your problem it has to be a problem
relating between hobbit and your machine (IO, disk, memory, privileges,
something).
quoted from Mike Livenspargar

On 10/10/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid> wrote:
 And now, just as magically as the CPU Load and Users and Processes graphs
started showing data, they have stopped again. All other trend graphs are
functioning. As far as I can tell all the data is coming over from the
client and it appears to be some issue with getting the data where it needs
to go to show up in the trend graphs, but I'm just guessing here.

Are there any thoughts on what I could look at to try to resolve this
problem?

Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
 Well, the trend graphs are showing data again. The server machine was
rebooted but it appears the graphs were showing data prior to the reboot. If
I had to guess I'd think that some process running on the client caused the
data parsing on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can examine that
might help me figure this out?

 I'll keep my eye on this to see what happens next.

Mike

-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent it, poorly.
--- Henry Spencer
list Mike Livenspargar · Wed, 10 Oct 2007 15:14:07 -0700 ·
As best I can tell all the possibilities you list below look
like they should present no problem to the process.
 
Since everything is working except for the two trend graphs
I noted, perhaps it would be best if I follow the data flow
for one of those graphs. I assume the data for the CPU Load
trend graph is pulled from the [uptime] section of the
client data report (I'm looking at what is shown when
clicking on the 'Client data available' link at the bottom
of the cpu status page). Where in the process can I put in
any hooks to verify that the data is moving through as
expected?
 
Thanks for all your help, and sorry for my lack of hobbit
knowledge.
Mike
user-4c45a83f15cb@xymon.invalid 10/10/2007 10:51 AM >>>
quoted from Josh Luthman
Have you tried checking to see if there are any quota or
disk issues?  If you were able to reboot the machine and get
your RRD activity back up while restarting the process did
not fix your problem it has to be a problem relating between
hobbit and your machine (IO, disk, memory, privileges,
something). 

On 10/10/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid>

wrote:And now, just as magically as the CPU Load and Users
quoted from Josh Luthman
and Processes graphs started showing data, they have stopped
again. All other trend graphs are functioning. As far as I
can tell all the data is coming over from the client and it
appears to be some issue with getting the data where it
needs to go to show up in the trend graphs, but I'm just
guessing here. 
 
Are there any thoughts on what I could look at to try to
resolve this problem?
 
Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
Well, the trend graphs are showing data again. The server
machine was rebooted but it appears the graphs were showing
data prior to the reboot. If I had to guess I'd think that
some process running on the client caused the data parsing
on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can
examine that might help me figure this out? 
 
 I'll keep my eye on this to see what happens next.
 
Mike


-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent
it, poorly. 
--- Henry Spencer
list Paul van Eldijk · Thu, 11 Oct 2007 11:49:05 +0200 ·
Mike Livenspargar schreef:
quoted from Mike Livenspargar
And now, just as magically as the CPU Load and Users and
Processes graphs started showing data, they have stopped
again. All other trend graphs are functioning. As far as I
can tell all the data is coming over from the client and it
appears to be some issue with getting the data where it
needs to go to show up in the trend graphs, but I'm just
guessing here.
exactly the same problem here: la, users and procs stopped reporting. No core dumps, no error messages, nothing. The data send from the client looks good, all data is present.

I've tried removing the rrds which weren't updated anymore but they are not recreated. Stopping/starting the client and/or hobbit server didn't resolve it either.

Hope someone knows a solution which does not involve rebooting the client, because that's not an option.

Paul
quoted from Mike Livenspargar
 Are there any thoughts on what I could look at to try to
resolve this problem?
 Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
Well, the trend graphs are showing data again. The server
machine was rebooted but it appears the graphs were showing
data prior to the reboot. If I had to guess I'd think that
some process running on the client caused the data parsing
on the server to go awry and once that process ended things
started working again. Is there any hobbit-saved data I can
examine that might help me figure this out?
  I'll keep my eye on this to see what happens next.
 Mike

list Mike Livenspargar · Thu, 11 Oct 2007 07:13:13 -0700 ·
And they're back! It seems that all that is needed to get these two trend graphs working again is 24 hours. In both cases where CPU load and users/processes trend graphs have stopped working they have started working again 24 hours after stopping. The 24-hour periods began at different times of the day * one at about 3 pm local, the other at about 5 am local.
 There were no reboots or restarts of server or client required in order to get things working again. I can't correlate any tasks scheduled to run on those machines that might have caused the graphs to start working again.  Does this offer any clues that will help us resolve this problem?
 Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/10/2007 3:14 PM >>>
quoted from Mike Livenspargar
As best I can tell all the possibilities you list below look like they should present no problem to the process.
 Since everything is working except for the two trend graphs I noted, perhaps it would be best if I follow the data flow for one of those graphs. I assume the data for the CPU Load trend graph is pulled from the [uptime] section of the client data report (I'm looking at what is shown when clicking on the 'Client data available' link at the bottom of the cpu status page). Where in the process can I put in any hooks to verify that the data is moving through as expected?
 Thanks for all your help, and sorry for my lack of hobbit knowledge.
Mike
user-4c45a83f15cb@xymon.invalid 10/10/2007 10:51 AM >>>
Have you tried checking to see if there are any quota or disk issues?  If you were able to reboot the machine and get your RRD activity back up while restarting the process did not fix your problem it has to be a problem relating between hobbit and your machine (IO, disk, memory, privileges, something). 
On 10/10/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid> wrote: And now, just as magically as the CPU Load and Users and Processes graphs started showing data, they have stopped again. All other trend graphs are functioning. As far as I can tell all the data is coming over from the client and it appears to be some issue with getting the data where it needs to go to show up in the trend graphs, but I'm just guessing here.  Are there any thoughts on what I could look at to try to resolve this problem?
 Thanks again,
Mike
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
Well, the trend graphs are showing data again. The server machine was rebooted but it appears the graphs were showing data prior to the reboot. If I had to guess I'd think that some process running on the client caused the data parsing on the server to go awry and once that process ended things started working again. Is there any hobbit-saved data I can examine that might help me figure this out?   I'll keep my eye on this to see what happens next.
 Mike


-- 
Josh Luthman
Office: XXX-XXX-XXXX
Direct: XXX-XXX-XXXX
XXXX Wayne St
Suite XXXX
Troy, OH XXXXX

Those who don't understand UNIX are condemned to reinvent it, poorly. --- Henry Spencer
list Tom L. Stewart · Thu, 11 Oct 2007 09:58:56 -0500 ·
I have the problem on a Solaris box, and my uptime is missing a field
(up xx days  and the who -b returns blank). This is because we do not
have an entry in wtmpx for the last boot record. The only way to fix the
issue to to reboot the system. I am not able to do that yet.  This is
all conjecture at this time, but the only thing that looked off in the
hobbitclient-sunos.sh page that is passed to Hobbit was the missing
field in uptime.
 
I can not get the box scheduled to reboot for another 2 months... so I
can't prove out my theory yet. If someone comes up with an answer,
please let everyone know. 
quoted from Mike Livenspargar
 
Tom 


	From: Mike Livenspargar [mailto:user-c6d9d5761955@xymon.invalid] 
	Sent: Wednesday, October 10, 2007 5:14 PM
	To: user-ae9b8668bcde@xymon.invalid
	Subject: Re: [hobbit] Trend Graphs Stopped Showing Data
	
	
	As best I can tell all the possibilities you list below look
like they should present no problem to the process.
	 
	Since everything is working except for the two trend graphs I
noted, perhaps it would be best if I follow the data flow for one of
those graphs. I assume the data for the CPU Load trend graph is pulled
from the [uptime] section of the client data report (I'm looking at what
is shown when clicking on the 'Client data available' link at the bottom
of the cpu status page). Where in the process can I put in any hooks to
verify that the data is moving through as expected?
	 
	Thanks for all your help, and sorry for my lack of hobbit
knowledge.
	Mike
	
user-4c45a83f15cb@xymon.invalid 10/10/2007 10:51 AM >>>
	
	Have you tried checking to see if there are any quota or disk
issues?  If you were able to reboot the machine and get your RRD
activity back up while restarting the process did not fix your problem
it has to be a problem relating between hobbit and your machine (IO,
disk, memory, privileges, something). 
	
	
	On 10/10/07, Mike Livenspargar <user-c6d9d5761955@xymon.invalid>
wrote: 

		And now, just as magically as the CPU Load and Users and
Processes graphs started showing data, they have stopped again. All
other trend graphs are functioning. As far as I can tell all the data is
coming over from the client and it appears to be some issue with getting
the data where it needs to go to show up in the trend graphs, but I'm
just guessing here. 
		 
		Are there any thoughts on what I could look at to try to
resolve this problem?
		 
		Thanks again,
		Mike
		
user-c6d9d5761955@xymon.invalid 10/9/2007 12:45 PM >>>
		
		
		Well, the trend graphs are showing data again. The
server machine was rebooted but it appears the graphs were showing data
prior to the reboot. If I had to guess I'd think that some process
running on the client caused the data parsing on the server to go awry
and once that process ended things started working again. Is there any
hobbit-saved data I can examine that might help me figure this out? 
		 
		 I'll keep my eye on this to see what happens next.
		 
		Mike
		
		
	-- 
	Josh Luthman
	Office: XXX-XXX-XXXX
	Direct: XXX-XXX-XXXX
	XXXX Wayne St
	Suite XXXX
	Troy, OH XXXXX
	
	Those who don't understand UNIX are condemned to reinvent it,
poorly. 
	--- Henry Spencer
list Paul van Eldijk · Fri, 12 Oct 2007 10:12:18 +0200 ·
Mike Livenspargar schreef:
quoted from Mike Livenspargar
And they're back! It seems that all that is needed to get these two trend graphs working again is 24 hours. In both cases where CPU load and users/processes trend graphs have stopped working they have started working again 24 hours after stopping. The 24-hour periods began at different times of the day * one at about 3 pm local, the other at about 5 am local.
 There were no reboots or restarts of server or client required in order to get things working again. I can't correlate any tasks scheduled to run on those machines that might have caused the graphs to start working again.  Does this offer any clues that will help us resolve this problem?
Exactly the same here! It cann't be coincidence that it resumes reporting after 24 hour. Anybody got a clue? Henrik?

greetings,
Paul
list Mike Livenspargar · Fri, 12 Oct 2007 06:58:33 -0700 ·
Paul:
 
In the interest of narrowing this problem down, what
platform are you on? In our case both client and server are
SLES v10 SP1.
 
Mike
user-86c97c6293b5@xymon.invalid 10/12/2007 1:12 AM >>>
quoted from Paul van Eldijk
Mike Livenspargar schreef:
And they're back! It seems that all that is needed to get
these two trend graphs working again is 24 hours. In both
cases where CPU load and users/processes trend graphs have
stopped working they have started working again 24 hours
after stopping. The 24-hour periods began at different times
of the day * one at about 3 pm local, the other at about 5
am local.
 
There were no reboots or restarts of server or client
required in order to get things working again. I can't
correlate any tasks scheduled to run on those machines that
might have caused the graphs to start working again. 
 
Does this offer any clues that will help us resolve this
problem?
Exactly the same here! It cann't be coincidence that it
resumes 
reporting after 24 hour. Anybody got a clue? Henrik?

greetings,
Paul
list Paul van Eldijk · Fri, 12 Oct 2007 16:25:12 +0200 ·
Mike Livenspargar schreef:
quoted from Mike Livenspargar
Paul:
 
In the interest of narrowing this problem down, what
platform are you on? In our case both client and server are
SLES v10 SP1.
Out hobbit server is an old box wit SuSE 9.3 (sic); old but running 
smoothly ;)
The client is running SuSE 10.0.

grtz,
Paul