nv-l
[Top] [All Lists]

RE: [nv-l] Another newbie Question - SNMP vs ICMP polling (setup)

To: <nv-l@lists.us.ibm.com>
Subject: RE: [nv-l] Another newbie Question - SNMP vs ICMP polling (setup)
From: "John Sobrinho" <johnsobrinho@rogers.com>
Date: Tue, 16 Mar 2004 18:06:20 -0500
Delivery-date: Tue, 16 Mar 2004 23:33:14 +0000
Envelope-to: nv-l-archive@lists.skills-1st.co.uk
Importance: Normal
In-reply-to: <OF356FBE45.56EAAE2B-ON87256E59.004AE436-85256E59.004C7E89@us.ibm.com>
Reply-to: nv-l@lists.us.ibm.com
Sender: owner-nv-l@lists.us.ibm.com
Thanks Jamie/Leslie and .. reamd,
 
I re-reviewd the NV 7 Admin guide and  between your comments and the guide, I got it all straightened up..
 
thanks all for you input. 
 
Cheers,
john
 
-----Original Message-----
From: owner-nv-l@lists.us.ibm.com [mailto:owner-nv-l@lists.us.ibm.com]On Behalf Of James Shanks
Sent: Tuesday, March 16, 2004 8:56 AM
To: nv-l@lists.us.ibm.com
Subject: Re: [nv-l] Another newbie Question - SNMP vs ICMP polling (setup)


Please don't call it NetView6000.  That terminology expired in 1998 when Solaris became the first non-IBM operating system supported by NetView.  I know it still appears internally in places, but the product is NetView for UNIX.

1.  Traps are an SNMP entity but your settings for SNMP polling have nothing to do with them.  trapd will accept and display whatever he gets, according to what is in trapd.conf, no matter what the source and no matter what community name is in the trap.  Traps from remote devices (Cisco routers for example), which have to be configured to send them to NetView, are directed to the default trap reception port, 162/udp.  Local traps, those from NetView itself, arrive via TCP using a UNIX  domain socket for the default port, normally 162/tcp.  Local traps initiated by NetView don't go out over the network and therefore cost nothing in terms of bandwidth.  Remote traps do have an associated cost so it is wise to configure your devices to send them sparingly.

2. The timing for polling is set in the xnmsnmpconf panel.  The standard polling methods are ICMP for status on a fairly short interval, and daily by SNMP for a configuration check.  You can make either of these happen at any time you choose with nmdemandpoll, which is called simply a "demand poll".  You can do that from the command line of by highlighting the node on the map and doing it from there.  What's in xnmsnmpconf are the standard intervals used by netmon.

3. There is no procedure to set up ICMP (ping) polling.  That's the default.  All you need do is set the status polling interval, time-out , and retries counts you'd prefer  in xnmsnmpconf.   If you want to have devices which are never to be status-checked by ping but rather by SNMP-only, then you put them in the netmon seed file preceded by the dollar sign, "$".  Then when the status polling intervals in xnmsnmpconf are reached, SNMP will be used for status instead of ping.  

Hope this helps

James Shanks
Level 3 Support  for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group



"John Sobrinho" <johnsobrinho@rogers.com>
Sent by: owner-nv-l@lists.us.ibm.com

03/15/2004 09:40 PM
Please respond to
nv-l

To
<nv-l@lists.us.ibm.com>
cc
Subject
[nv-l] Another newbie Question - SNMP vs ICMP polling (setup)





 
 
 
Hmmm ......

Can someone set me straight.

My objective : use Netview mapping and  NV6000 traps to work.
 
We are already using netview to accept traps and sending tec event out for specific traps. I know I am recieving NV6000 alerts, but is t being done by SNMP or ICMP.
 
If I want SNMP polling is it just a matter of adding $ in front of my entries in the seed file and setting up the snmp configuration panels ? I thought the snmp config panel also controls the ping interval time ?

What is the procedure to set up ICMP and procedure to set up SNMP polling ?

 
thanks
 
John
 
 
<Prev in Thread] Current Thread [Next in Thread>

Archive operated by Skills 1st Ltd

See also: The NetView Web