Xymon Mailing List Archive search

Two issues - new user

7 messages in this thread

list Brian O'Mahony · Wed, 2 May 2007 17:47:18 +0100 ·

-----Original Message-----
From: Brian O'Mahony 
Sent: 02 May 2007 17:36
To: user-1bd859b2eb47@xymon.invalid
Subject: Two issues - new user

Ive just taken over managing the unix servers here at work - they are a
combination of AIX, HPUX, and solaris.

Here are the three issues I have stumbled across, and don't know enough
about this program to resolve:

#1
One of our HPUX PA-RISC machines has a red smiley face next to ports.
This server was previously configured to run hobbit. Here is the error :
Weblogic (found 0, req. 1 or more)

I thought someone may have set it up to watch the Weblogic port, so I
removed the older version of hobbit, and put on the latest stable
release. Problem is the face is still red. How do I remove this?

#2
An AIX server (5.3) every now and again the memory page goes red. It
says its using 4billion % of the memory. While this seems great
efficiency for such a small box how can I prevent this / why is it
happening?

#3 
Some of the servers are yellow for memory, and when expanded it points
to vm/swap space. The servers sometimes go above 80% (which I can see
from the configuration report). However they just stay yellow even when
they return below the threshold. How do I stop these from staying
yellow, or at least when they do give a yellow alert, that I can just
turn it back to green after noting it?

Thx

Brian


The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorized. If you are not the intended recipient, any disclosure,
copying, distribution or any action taken or omitted to be taken in reliance
on it, is prohibited and may be unlawful. If you are not the intended
addressee please contact the sender and dispose of this e-mail. Thank you.
list Tom Georgoulias · Wed, 02 May 2007 15:25:47 -0400 ·
quoted from Brian O'Mahony
Brian O'Mahony wrote:
#1
One of our HPUX PA-RISC machines has a red smiley face next to ports.
This server was previously configured to run hobbit. Here is the error :
Weblogic (found 0, req. 1 or more)

I thought someone may have set it up to watch the Weblogic port, so I
removed the older version of hobbit, and put on the latest stable
release. Problem is the face is still red. How do I remove this?
That's what it sounds like to me.  To remove the port test for Weblogic, open the the hobbit-client.cfg file on the Hobbit server, find the client config that corresponds to the hostname of the server showing up as red, and delete the PORT line for Weblogic.  It will probably look something like this:

PORT LOCAL=xxxxxxxx TEXT=Weblogic
quoted from Brian O'Mahony
#2
An AIX server (5.3) every now and again the memory page goes red. It
says its using 4billion % of the memory. While this seems great
efficiency for such a small box how can I prevent this / why is it
happening?
Do you get weird values when the memory test status is green or yellow?   Can you post an example?
quoted from Brian O'Mahony
#3 Some of the servers are yellow for memory, and when expanded it points
to vm/swap space. The servers sometimes go above 80% (which I can see
from the configuration report). However they just stay yellow even when
they return below the threshold. How do I stop these from staying
yellow, or at least when they do give a yellow alert, that I can just
turn it back to green after noting it?
You cannot remove the yellow condition or change the color to green unless the system itself is no longer reporting memory levels that exceed the yellow alert threshold.  Either set the yellow alert levels higher or use 101, which prevents it from alerting on any value from 0-100.
-- 
Tom Georgoulias
Systems Engineer
McClatchy Interactive
user-6a0b8b0f0ae1@xymon.invalid
list Henrik Størner · Wed, 2 May 2007 22:39:11 +0200 ·
quoted from Brian O'Mahony
On Wed, May 02, 2007 at 05:47:18PM +0100, Brian O'Mahony wrote:
#1
One of our HPUX PA-RISC machines has a red smiley face next to ports.
This server was previously configured to run hobbit. Here is the error :
Weblogic (found 0, req. 1 or more)
"next to ports" - do you mean in the "procs" column ? It sounds like it
is. In that case, you'll find there is a setting in the
"hobbit-clients.cfg" file where this host has configured a check for a
process called "WebLogic". Remove that check, and the red smiley should
turn green.
quoted from Tom Georgoulias
#2
An AIX server (5.3) every now and again the memory page goes red. It
says its using 4billion % of the memory. While this seems great
efficiency for such a small box how can I prevent this / why is it
happening?
Sounds like an integer overflow in how the percentage is calculated.
Could you forward me the detailed memory status when this happens ?
You can find one of them if you click on the "History" button for the
memory status, and then click on one of the red occurrences.
quoted from Tom Georgoulias
#3 
Some of the servers are yellow for memory, and when expanded it points
to vm/swap space. The servers sometimes go above 80% (which I can see
from the configuration report). However they just stay yellow even when
they return below the threshold. How do I stop these from staying
yellow, or at least when they do give a yellow alert, that I can just
turn it back to green after noting it?
The thresholds for when they go yellow (or red) are in the
hobbit-clients.cfg file (on the Hobbit server). You can change them there.


Regards,
Henrik
list Brian O'Mahony · Thu, 3 May 2007 08:57:46 +0100 ·
quoted from Henrik Størner
#2
An AIX server (5.3) every now and again the memory page goes red. It
says its using 4billion % of the memory. While this seems great
efficiency for such a small box how can I prevent this / why is it
happening?
Sounds like an integer overflow in how the percentage is calculated.
Could you forward me the detailed memory status when this happens ?
You can find one of them if you click on the "History" button for the
memory status, and then click on one of the red occurrences.

Thanx

#1 & #3 are resolved.

The weird memory issue has happened again this morning:

Thu May 3 08:54:21 BST 2007 - Memory CRITICAL

   Memory              Used       Total  Percentage
red Physical     4294967205M       4096M 4294967294%
green Swap                 87M       8704M          0%

It lasted for about 5 mins then went back to normal.

B
quoted from Brian O'Mahony


The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorized. If you are not the intended recipient, any disclosure,
copying, distribution or any action taken or omitted to be taken in reliance
on it, is prohibited and may be unlawful. If you are not the intended
addressee please contact the sender and dispose of this e-mail. Thank you.
list Henrik Størner · Thu, 3 May 2007 11:11:12 +0200 ·
quoted from Brian O'Mahony
On Thu, May 03, 2007 at 08:57:46AM +0100, Brian O'Mahony wrote:
The weird memory issue has happened again this morning:

Thu May 3 08:54:21 BST 2007 - Memory CRITICAL

   Memory              Used       Total  Percentage
red Physical     4294967205M       4096M 4294967294%
green Swap                 87M       8704M          0%

It lasted for about 5 mins then went back to normal.
Could you find this status message in the "History" of the events, and
click on the "Client data" link at the bottom of the message ? I would
like to see the raw data behind this, they are in the "[realmem]" and
"[freemem]" sections of that data.


Regards,
Henrik
list Brian O'Mahony · Thu, 3 May 2007 10:20:41 +0100 ·
Im not sure what you are asking here :
I click on the mem section, go to history, and then go down to "Last 50
entries" click on the red flashing icon, and it takes me to the event.
However there is no client data link there.

It happened again (duration 50 mins) so next time it happens I can just
grab it when it occurs (there is a link to client data available when I
click on the memory section of the server)

B
quoted from Brian O'Mahony


-----Original Message-----
From: Henrik Stoerner [mailto:user-ce4a2c883f75@xymon.invalid] 
Sent: 03 May 2007 10:11
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] FW: Two issues - new user

Could you find this status message in the "History" of the events, and
click on the "Client data" link at the bottom of the message ? I would
like to see the raw data behind this, they are in the "[realmem]" and
"[freemem]" sections of that data.


Regards,
Henrik


The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorized. If you are not the intended recipient, any disclosure,
copying, distribution or any action taken or omitted to be taken in reliance
on it, is prohibited and may be unlawful. If you are not the intended
addressee please contact the sender and dispose of this e-mail. Thank you.
list Brian O'Mahony · Thu, 3 May 2007 13:53:22 +0100 ·
quoted from Brian O'Mahony

-----Original Message-----
From: Brian O'Mahony 
Sent: 03 May 2007 13:13
To: user-ae9b8668bcde@xymon.invalid
Subject: RE: [hobbit] FW: Two issues - new user

Just caught one: (as you can see the freemem section is empty)

Thu May 3 13:14:34 BST 2007 - Memory CRITICAL

   Memory              Used       Total  Percentage
red Physical     4294965248M       4096M 4294967246%
green Swap                 87M       8704M          0%

----

client morticia,curamsoftware,com.aix aix
[date]
Thu May  3 13:14:34 BST 2007
[uname]
AIX morticia 3 5 005CFACA4C00
[uptime]
  01:14PM   up 2 days,   1:23,  1 user,  load average: 0.01, 0.05, 0.08
[who]
qaconfig    pts/1       May  2 09:54     (eschowe.curamsoftware.com)
[df]
Filesystem    1024-blocks      Used      Free %Used Mounted on
/dev/hd4          1048576    302576    746000   29% /
/dev/hd2         10485760   4227072   6258688   41% /usr
/dev/hd9var       2097152    398628   1698524   20% /var
/dev/hd3          3670016    543628   3126388   15% /tmp
/dev/hd1         10485760   6087340   4398420   59% /home
/proc                   -         -         -    -  /proc
/dev/hd10opt      6029312    108368   5920944    2% /opt
/dev/fslv00      10485760   7875996   2609764   76% /oracle9
/dev/fslv01      10485760   5945936   4539824   57% /oracle10
[realmem]
realmem 4194304 Amount of usable physical memory in Kbytes False
[freemem]
[swap]
Total Paging Space   Percent Used
      8704MB               1%
[ifconfig]
en0:
flags=5e080863,c0<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROU
PRT,64BIT,CHECKSUM_OFFLOAD,PSEG,CHAIN>
	inet 172.16.165.143 netmask 0xfffff000 broadcast 172.16.175.255
	 tcp_sendspace 65536
lo0:
flags=e08084b<UP,BROADCAST,LOOPBACK,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64
BIT>
	inet 127.0.0.1 netmask 0xff000000 broadcast 127.255.255.255
	inet6 ::1/0
	 tcp_sendspace 131072 tcp_recvspace 131072 rfc1323 1
[route]


(I had to snip the rest cos the mail server wouldn't allow it - if there
is anything else you need to see let me know)
quoted from Brian O'Mahony


-----Original Message-----
From: Brian O'Mahony [mailto:user-9ed4e9656005@xymon.invalid] 
Sent: 03 May 2007 10:21
To: user-ae9b8668bcde@xymon.invalid
Subject: RE: [hobbit] FW: Two issues - new user

Im not sure what you are asking here :
I click on the mem section, go to history, and then go down to "Last 50
entries" click on the red flashing icon, and it takes me to the event.
However there is no client data link there.

It happened again (duration 50 mins) so next time it happens I can just
grab it when it occurs (there is a link to client data available when I
click on the memory section of the server)

B


The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorized. If you are not the intended recipient, any disclosure,
copying, distribution or any action taken or omitted to be taken in reliance
on it, is prohibited and may be unlawful. If you are not the intended
addressee please contact the sender and dispose of this e-mail. Thank you.