| Author |
Message |
John Mrav
Guest
|
Posted:
Thu Jan 27, 2005 6:48 am Post subject:
Agentless monitoring generating high Netbios-IP traffic |
|
|
Hi,
I am successfully monitoring a number of NT4 SP6a agentlessly, but am
noticing what I consider extremely high bandwidth utilisation. Each server is
generating approx. 80kbps of Netbios-IP (port 139) traffic back to the MOM
server.
Is this expected or does it look like there is a problem here?
Any ideas?
Thankss,
John |
|
| Back to top |
|
 |
Kelly Morris [MSFT]
Guest
|
Posted:
Mon Jan 31, 2005 6:47 am Post subject:
Re: Agentless monitoring generating high Netbios-IP traffic |
|
|
Good Morning John,
I can't give you hard numbers but, this appears to be expected. When
you monitor a server agentlessly, the MOM server acts as the agent for that
machine. It must scan the registry, read the log files, and retrieve
performance counter data all remotely. The act of opening a machine's
registry for remote reading uses a fair amount of bandwidth.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:FF42383E-7D22-4FEB-84CF-12F660C17D6E@microsoft.com...
| Quote: | Hi,
I am successfully monitoring a number of NT4 SP6a agentlessly, but am
noticing what I consider extremely high bandwidth utilisation. Each server
is
generating approx. 80kbps of Netbios-IP (port 139) traffic back to the MOM
server.
Is this expected or does it look like there is a problem here?
Any ideas?
Thankss,
John |
|
|
| Back to top |
|
 |
John Mrav
Guest
|
Posted:
Tue Feb 01, 2005 6:49 am Post subject:
Re: Agentless monitoring generating high Netbios-IP traffic |
|
|
Hey Kelly,
Can you please tell me, is it possible to minimise the number of
resources/counters/logs that are monitored on agentless servers without
affecting servers with agents installed?
With these particular servers I'd be happy to know that they are up, their
aren't full, the CPU isn't high and there is memory available :)
Thanks,
John
"Kelly Morris [MSFT]" wrote:
| Quote: | Good Morning John,
I can't give you hard numbers but, this appears to be expected. When
you monitor a server agentlessly, the MOM server acts as the agent for that
machine. It must scan the registry, read the log files, and retrieve
performance counter data all remotely. The act of opening a machine's
registry for remote reading uses a fair amount of bandwidth.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:FF42383E-7D22-4FEB-84CF-12F660C17D6E@microsoft.com...
Hi,
I am successfully monitoring a number of NT4 SP6a agentlessly, but am
noticing what I consider extremely high bandwidth utilisation. Each server
is
generating approx. 80kbps of Netbios-IP (port 139) traffic back to the MOM
server.
Is this expected or does it look like there is a problem here?
Any ideas?
Thankss,
John
|
|
|
| Back to top |
|
 |
Kelly Morris [MSFT]
Guest
|
Posted:
Wed Feb 02, 2005 2:16 am Post subject:
Re: Agentless monitoring generating high Netbios-IP traffic |
|
|
It may be possible to exclude those computers from specific computer groups.
In the operations console, select the computers view. Select one of the
agentless machines and then look in the details pane. You should see a tab
for Computer Groups. This will tell you to which computer groups the
selected computer belongs. In the administration console, select the
computer group specified in the operations console computers view. View the
properties for this group and select the excluded computers tab. You should
be able to explicitly add the agentless machine to the exclusions list.
This would prevent rules associated with those computer groups from running
against the agentless machines. Do this for all the computer groups for
which you don't want the associate rules running on the agentless machines.
You could then create a new computer group for these machines, create a new
management pack with just the rules you want to run, and then associate the
new MP with the new computer group.
This should minimize the amount of traffic between these machines and the
MOM server. Understand though, agentless management is by definition
network intensive, network traffic will depend more on the number of
agentless machines managed then the number of rules per agentless machine.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:760C195B-5AAF-479B-B37E-24F69055653B@microsoft.com...
| Quote: | Hey Kelly,
Can you please tell me, is it possible to minimise the number of
resources/counters/logs that are monitored on agentless servers without
affecting servers with agents installed?
With these particular servers I'd be happy to know that they are up, their
aren't full, the CPU isn't high and there is memory available :)
Thanks,
John
"Kelly Morris [MSFT]" wrote:
Good Morning John,
I can't give you hard numbers but, this appears to be expected. When
you monitor a server agentlessly, the MOM server acts as the agent for
that
machine. It must scan the registry, read the log files, and retrieve
performance counter data all remotely. The act of opening a machine's
registry for remote reading uses a fair amount of bandwidth.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:FF42383E-7D22-4FEB-84CF-12F660C17D6E@microsoft.com...
Hi,
I am successfully monitoring a number of NT4 SP6a agentlessly, but am
noticing what I consider extremely high bandwidth utilisation. Each
server
is
generating approx. 80kbps of Netbios-IP (port 139) traffic back to the
MOM
server.
Is this expected or does it look like there is a problem here?
Any ideas?
Thankss,
John
|
|
|
| Back to top |
|
 |
John Mrav
Guest
|
Posted:
Wed Feb 02, 2005 1:19 pm Post subject:
Re: Agentless monitoring generating high Netbios-IP traffic |
|
|
Thanks Kelly, I'll try this.
John :^)
"Kelly Morris [MSFT]" wrote:
| Quote: | It may be possible to exclude those computers from specific computer groups.
In the operations console, select the computers view. Select one of the
agentless machines and then look in the details pane. You should see a tab
for Computer Groups. This will tell you to which computer groups the
selected computer belongs. In the administration console, select the
computer group specified in the operations console computers view. View the
properties for this group and select the excluded computers tab. You should
be able to explicitly add the agentless machine to the exclusions list.
This would prevent rules associated with those computer groups from running
against the agentless machines. Do this for all the computer groups for
which you don't want the associate rules running on the agentless machines.
You could then create a new computer group for these machines, create a new
management pack with just the rules you want to run, and then associate the
new MP with the new computer group.
This should minimize the amount of traffic between these machines and the
MOM server. Understand though, agentless management is by definition
network intensive, network traffic will depend more on the number of
agentless machines managed then the number of rules per agentless machine.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:760C195B-5AAF-479B-B37E-24F69055653B@microsoft.com...
Hey Kelly,
Can you please tell me, is it possible to minimise the number of
resources/counters/logs that are monitored on agentless servers without
affecting servers with agents installed?
With these particular servers I'd be happy to know that they are up, their
aren't full, the CPU isn't high and there is memory available :)
Thanks,
John
"Kelly Morris [MSFT]" wrote:
Good Morning John,
I can't give you hard numbers but, this appears to be expected. When
you monitor a server agentlessly, the MOM server acts as the agent for
that
machine. It must scan the registry, read the log files, and retrieve
performance counter data all remotely. The act of opening a machine's
registry for remote reading uses a fair amount of bandwidth.
--
Kelly Morris [MSFT]
This post is provided "AS IS" with no warranties, and confers no rights.
"John Mrav" <JohnMrav@discussions.microsoft.com> wrote in message
news:FF42383E-7D22-4FEB-84CF-12F660C17D6E@microsoft.com...
Hi,
I am successfully monitoring a number of NT4 SP6a agentlessly, but am
noticing what I consider extremely high bandwidth utilisation. Each
server
is
generating approx. 80kbps of Netbios-IP (port 139) traffic back to the
MOM
server.
Is this expected or does it look like there is a problem here?
Any ideas?
Thankss,
John
|
|
|
| Back to top |
|
 |
|
|
|
|