SNMP and netware 4.11 1

[Thnoss]

Can anyone assist with setting up SNMP on my novell server. Also is there anything special that needs to be done to setup tcp/ip on a netware 4.11?

Thanks guys

 

[sobak]

What exactly are you attempting to do? Monitor your server via SNMP or send traps from the server to a monitoring box?

I know you can setup SNMP to send traps via the INETCFG nlm, I didn't like to load all my settings in the INETCFG file so I opted to set it up manually. I'll have to find the SNMP line for loading it via autoexec.ncf.

If you're attempting to monitor the system via an SNMP client then you need a little file called servinst.nlm. I believe if you search for it on Novell's site then you should be able to find it. Keep in mind that it helps to have the Novell MIBs (something I found out the hard way)loaded on your SNMP client.



david e
*end users are just like computers, some you can work with...others just need a simple reBOOTing to fix their problems.*

 

[Thnoss]

I would like to use SNMP to monitor the server, excuse my ingorance but what do you mean by MIBs?
Ill have a look for this file you mentioned. Thanks for the help.

 

[jjrowan]

I was catching up on the Tek-Tips Novell forum and saw your June 2003 post concerning monitoring Netware servers. Did you implement something for monitoring the servers? Querying via SNMP is one way of monitoring but how often are you going to send an SNMP packet? SNMP is User Datagram Protocol (UDP) so there is no acknowledgement other than the response. If you're on a busy network UDP can get dropped. Also, if your network is busy do you want to increase that traffic by adding an SNMP? NetWare broadcasts IPX and SPX information every minute IPX then 30 seconds later SPX (if I remember correctly). You could listen to this information to determine whether your server is alive. I use to do this for servers and then for dedicated DOS (does anyone still use DOS?) PCs. If a server's SAP packet wasn't seen for 5 consecutive minutes I would alert the admin that the server was down. On the DOS workstation side of things the workstation has a diagnostic stack which can be queried to determine if it's actually processing. If the number of packets in a given time frame isn't within a percentage of the norm I would send a reboot packet to that workstation. If it still didn't come up to speed then I'd alert the admin. It's been years since I've done Netware (Linux now), I don't know if this is still the case with v 5 and 6.