Larry,
If the device is in an "Unreachable" region, it will not be polled
and the status will not be changed.
Notice this message:
14:36:08 Interface 10.10.10.20(sc0) is in
UNREACHABLE region
Paul
Larry Fagan wrote:
I'm sorry guys if i'm confusing you all.. But james it
is true..
I have a switch which is turned off and i have it as
green in map. But the ping/SNMP connectvity fails.
When i do demand poll from command line.. Out put is
below:
nmdemandpoll NETVC3111.uscom
14:35:53 ***** Starting demand poll of node
NETC3111.us.com *****
14:35:53 Interface 10.10.10.20(sc0) (currently up)
14:36:08 Interface 10.10.10.20(sc0) is in
UNREACHABLE region
14:36:08 Current polling parameters
14:36:08 scheduled configuration check at 03/12/05
10:30:51
14:36:08 scheduled new node poll at 03/10/05
14:48:42
14:36:08 auto-adjusted polling interval is 1620
seconds
14:36:08 Verify node name
14:36:08 node name verified to be NETVC3111.us.com
14:36:08 ***** End of demand poll for node
NETVC3111.us.com
..
What does this means by interface currently up..?
Switch is down... i can't ping from netview server as
well..
I did netmon -a 12 and then netmon -M 10 but first -a
option said in 20 sec but later -M 10 did not show up
in the netmon.trace at all..
Also i did delete this switch from Map and re
-discovered today but still the problem exists.. What
can i do now..?
Many Many Thanks,
Larry
--- James Shanks <jshanks@us.ibm.com> wrote:
Larry,
I'm sure that I speak for others when I say that
most of us are as confused
as we can be about your description here.
As I understand it, you have switch which is down
but still it shows as
green on the map.
OK, what happens when you select it on the map, pull
down Test --> Demand
Poll? A little window should open to show you the
results of that demand
poll. If the thing is really down, netmon should
send the down trap, and
ipmap should turn it red. You can also do demand
poll from the command
line, if you want to capture the output. The
command is
/usr/OV/bin/nmdemandpoll <fully
qualified node name>
Just re-direct the output to a file,
/usr/OV/bin/nmdemandpoll <fully
qualified node name> >
nmdemandpoll.out
On your second question, about "netmon -a 12", I'm
confused as well. How
do you know that netmon is not polling this thing on
the stated interval?
Are you running a full netmon trace? That's the
only way to tell. To
do that you would issue
/usr/OV/bin/netmon -M <trace_mask>
where the trace mask is a numeric value (you can see
the possible values in
the man page for netmon, "man netmon").
But presumably, this switch is being polled by ICMP
(ping), so you could
just issue
/usr/OV/bin/netmon -M 2
to see those pings and their replies. If the thing
is pooled only by SNMP,
'-M 8' would show you those, and "netmon -M 10"
would show both ICMP and
SNMP requests and replies. All this gets written to
the netmon trace. To
stop the tracing, issue "netmon -M 0". To run a
full trace of absolutely
everything netmon is doing, issue "netmon -M -1". I
don't recommend you
trace it forever -- lots of overhead there - but if
you want to do that you
can set an option on the start up of the daemon
using serversetup.
The idea is that if "netmon -a 12" shows your switch
scheduled for a poll
in 88 seconds, then you'd issue one of those "netmon
-M" commands, and
after 88 seconds, go look at the netmon.trace file
to see what happened.
If you cannot interpret what you see, then a call to
Support is in order so
that a Level 2 specialist can help you figure it
out.
James Shanks
Level 3 Support for Tivoli NetView for UNIX and
Windows
Tivoli Software / IBM Software Group
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
|