Hi Sean,
You should reference the TME 10 NetView Mid-Level Manager User's Guide and the
netmon man page for details on how to set up the MLM if the following doesn't
help.
Your MLM is sending Node Up traps to NetView. That will happen when the MLM
is restarted, or when it detects the node is up after it has found it down.
I am not sure of the internals but I believe that netmon looks for these MLM
traps and sets the status of the node if it has changed. If netmon does
change the status, then you will see a NetView Node Up trap accompanying the
"MLM detected Node Up" trap.
If the system is configured correctly, NetView will assign nodes to the MLMs.
If it does, it should update the status polling intervals, and retries for
those nodes. (There was a defect that would cause NetView not to update these
intervals properly if there are nodes with different values for polling
interval and retries but I think it is fixed in 5.1.1.)
If you want to know what netmon actualy thinks it has assigned to the MLMs,
you can run the command "netmon -a 102" and netmon will append it's list of
mlms and managed interfaces to the /usr/OV/log/netmon.trace file. The "-a 51"
flag will dump the list of nodes and interfaces managed by mlms, and the "-a
50" will cause netmon to create collections for each MLM that will contain the
nodes managed by each. If netmon has not assigned any nodes or interfaces to
the MLM's, then something is misconfigured and it of course will not be able
to change the status polling interval. The most common misconfiguration would
be the /etc/snmpd.conf file on the MLM. If NetView cannot do an snmpset to
the MLM, it won't be able to properly assign nodes to the MLM.
There is also a config file used by the MLM for SNMP timeouts and community
names for snmpgets and snmpsets. It is in the same format as the file
maintained by xnmsnmpconf /usr/OV/conf/ovsnmp.conf (although the actual data
used is stored in a database in the directory /usr/OV/conf/ovsnmp.conf_db).
So you could either:
1. copy this file to the MLM
2. run xnmsnmpconf -export filename, and copy the resultant file "filename"
to the MLM renaming it to ovsnmp.conf.
3. NFS export /usr/OV/conf from the NetView server, and mount it as
/usr/OV/conf on the MLM.
The MLM looks for the ovsnmp.conf file first in /usr/OV/conf/ directory, and
then in /var/adm/smv2/ directory.
Sean Davidson wrote:
> I am getting quire a few of "MLM detected node up events" from netmon. I'm
> running NV 5.1.1, Tivoli 3.6.1, on AIX 4.2.1. My Netview Server is also the
> TMR and I have two MLM's.
>
> What causes these events? I've been increasing the timeout in xnmsnmpconf,
> but I'm not sure this is the right place. Here are the parms for netmon on
> my Netview Server:
>
> /usr/OV/bin/netmon -P -I -S -s/usr/OV/conf/netmon.seed -u
> -G/usr/OV/conf/mlm.seed -z -L
>
> /usr/OV/conf/mlm.seed just has my MLMs in it.
>
> /usr/OV/conf/netmon.seed has the MLMs and some !@ oid statements to only
> discover SNMP oids I want.
>
> I think I'm just looking in the wrong spot.
>
> Thanks! Sean x6889
--
Ray Schafer | schafer@tkg.com
The Kernel Group | Distributed Systems Management
http://www.tkg.com
|