Maxime,
You don't need to use the snmp proxy if your MLM can ping it. I thought
you had a network that both the MLM and NetView could not ping.
Do you have a lot of MLMs? A lot of nodes being managed by MLMs? What I
saw a couple of years ago was that netmon would have to start and fully
initialize before it started to distribute status polling to the specific
MLMs. Until then, netmon was doing the status polling and the interfaces
would be red if NetView cannot ping it. If netmon was restarted, this
would happen all over again.
I have also seen in the past netmon take a lot longer time to initialize if
you are only using /etc/hosts or only have a subset of managed interfaces
in your DNS. There are "netmon -a" commands that let you see what
interfaces have been distributed to MLMs. I would look that up for you,
but I am away from everything including documentation. I have done crazy
things like distribute 2000 interfaces for polling by an MLM, then came to
realize it took a long time for NetView to configure the MLM to poll that
many. The MLM was happy and could do the work, it was the configuration
that took too long for my customer.
My recommendation is to only use MLM's where you really need to. Don't use
MLMs to distribute polling unless you must do so for the physical network.
Kind regards,
Stephen Hochstetler shochste@us.ibm.com
International Technical Support Organization - Austin
Office - 512-436-8564 FAX - 512-436-8701
ITSO redbooks at http://www.redbooks.ibm.com
|