Navisphere Agent not running on Windows 2003 Server

From the Application Event Log, I find that it mentions that "No entries in the device list". I just wonder what the device list it means ?

Event Type: Information
Event Source: Navisphere Agent
Event Category: None
Event ID: 1
Date: 2/01/2008
Time: 2:22:44 PM
User: N/A
Computer: NEW_SERVER
Description:
No entries in the device list. This is normal if the host is attached only to FC4700 arrays.

Thanks

Couple of things to check:

1. If there are multiple NICs on the servers ensure that the NIC with the IP that communicates with the array is listed first in the binding order. This can cause the errors you describe if the server is registering an IP with the array that the array can't access.

2. Ensure that the agent.config file for Navisphere has the IPs of both Storage Processors listed in the correct format: user system@w.x.y.z

3. Check the connectivity status on the array and ensure that the agent is logged in AND registered.

I'm having similar issues.

My servers see the luns fine and everything apears to work, but I get something like:

[ Unknown; Fibre, Hoste Agent not reachable]

If you do have a multi-homed server you should also create ¿agentid.txt¿ file.

 Open up a new Notepad
 On the first line enter the name of the server (as you would like it to appear in Navisphere
 On the 2nd line enter the IP address of the server used for the main communication to the outside world
 Be sure to add a CRLF ('enter') after the IP address
 Save the file with the name: ¿agentid.txt¿ in the directory ¿ :\Program Files\EMC\Navisphere Agent¿,
 Close Notepad

Normally Navisphere Agent will pick-up this IP address, but for systems with multiple IP addresses (e.g. Clusters), it is not guaranteed that Navisphere will select the proper IP address (this is a Windows issue). By creating this file the proper IP address will be used by Navisphere.

I hope this helps
Carl

That step is not necessary as long as the binding order of the NICs is correct. My infrastructure is primarily Windows clusters on Clariion storage and I have never had a problem as long as the public NIC was at the top of the binging order. This is also a Microsoft best practice for clusters, so as long as cluster build guides are used it shouldn't be an issue.

Make sure that you don't have a firewall between the array and the hosts, if you do, you'll need to open TCP port 6389 bi-directional.

Also see Primus emc66921 about the multiple NIC issue and make sure that the correct entries are in the agent.config file:

user system@IP_Address_SPA
user system@IP_Address_SPA

regards,

glen kelley

Dear Aran,

Thank you for your advice.

We do have included both SP IP addresses in the agent.config file.

We find that one of the HBA is connected to the SAN but the other one is not.

In this way, we guess that there is problem between one HBA and Switch B. Although the lights on both HBA and Brocade Switch seem to be OK, we re-enable the Switch configuration.

After doing that, the connection is OK now.

Thanks