Make sure you are on the newest version AND make sure that your 
/etc/init.d/init.snmpdx start command includes the "-f 0" parameter

The problem is anytime the Solaris agent (which if you haven't read Pauls 
comment or my numerous rants on the subject) receives an "invalid" community 
string it passes the community string "public" down to the subagents which 
(without the -f 0) die.

-----Original Message-----
From: owner-nv-l AT lists.us.ibm DOT com [mailto:owner-nv-l AT lists.us.ibm DOT 
com]On
Behalf Of Paul
Sent: Friday, March 19, 2004 10:46 AM
To: nv-l AT lists.us.ibm DOT com
Subject: Re: [nv-l] Sun snmp daemon and netview


Fawad,
Solaris's snmp agent is a piece of junk. Restart it on those servers:

/etc/init.d/init.snmpdx stop
/etc/init.d/init.snmpdx start

And you should start getting responses again.

Paul


Qureshi, Fawad wrote:

> NV 7.1.3 / AIX 5.1
>
>  
>
> Has anyone experienced any problems using Sun's snmp daemon and 
> Netview? Especially with community names?
>
>  
>
> We have several Sun servers that we have been monitoring fine till few 
> days back. These servers do not respond to snmpwalk or any snmp 
> queries anymore. SNMP times out. Possible cause could be incorrect 
> community name, but I have checked and all looks correct on both 
> sides. I am able nmdemandpoll, TCP/SNMP connectivity etc. just fine, 
> except SNMP requests time out.
>
>  
>
> I have done xnmsnmpconf -resolve and result is correct. Forcing 
> community name with snmpwalk -c still results in time out due to bad 
> community.
>
>  
>
> "Bad Community name" alerts are also being logged in Sun's syslog files.
>
>  
>
> I have seen this before as well, and always with Sun servers. Any 
> ideas? Thanks.
>
>  
>
> **Cheers **
>
> ** **
>
> **Fawad Qureshi**
>
>  
>