DL380G3 Clustered Package Problem

[John Calder_1]

I have a DL380G3 Clustered Package. I am running Groupwise 6.5 on top of Netware 5.5. I ahve had the system for 3 months running perfect, clustered post offices. I have all my users divided into two post offices. Each server is running one post office (active/active configuration). If one server dies all the services switch over to the other box and it runs both post offices. Last week one machine started to abend and the post office would switch of by itself (which it should do). Problem is both servers have started having the same problem. Also if you go to either server when they are running and ping the other box either via the private 192 network or the public network the ping works fine for about 20-50 packets and they the send count keeps going up and the recieve stops. After the system has sent 3-5 more packets then received the system freezes for about 10 seconds and they all the received packets come in and the system works fine for about a minute or so. My theory is that this pause is cuasing he servers to abend and fail over. When the server stops like this I watch the activity light on the arrary and the drive light goes crazy, once the server starts going again the drive light slows do to a blink a second or two. I have replaced the NIC cards in both server and still same problem. I have updated the NIC drivers and same problem. I have another theory that the controller in the array is going bad and that is causing some problems. Has anyone seen this problem or have any idea what I could look at? Please send me an e-mail if you have questions or answers to 'damaddog@doglover.com'.

Thanks!

[Doug de Werd]

Got a few questions for you to help figure this out:

- Version of NetWare (5.5 is not a NetWare version, there's 5.1, 6 and 6.5) This is important since NetWare 5.1 had a stand alone cluster product and NetWare 6.x includes clustering with the OS
- Any Novell Support Packs applied? Version number?
- Any HP ProLiant Support Pack's applied? Version number?
- Has the latest ROM been applied to the servers? ROM version and date.
- Are you running multiple processors in the 380's (are you running CPQACPI.PSM or CPQMPK.PSM) Version number for the running module?
- What is the version of the Q57.LAN driver?
- What is the version of CPQSHD.CDM?

Thanks,
Doug