02-03-2004 05:57 AM - last edited on 05-22-2013 07:51 PM by maikoro
I'm currently in the process of specifying/purchasing a 2-node DL380+MSA500 cluster to replace our existing standalone LH3000s. This is my first exposure to clusters so please excuse the possibly stupid questions.
Looking at the MSA500 manual it states the power-up sequence should be the MSA500 and then the connected servers.
Our system will be UPS protected for controlled shutdown, but what happens when the mains power comes back, how can we ensure the system will startup correctly with no user intervention? (will the servers hang or wait for the SAS to startup?).
Do you recommend having a single UPS powering the servers+SAS? Assuming something like an APC UPS what is the best way of performing a controlled shutdown (I assume pick one server as RS232/UPS connected and put in a big time delay after it has shutdown)? How is the oether clustered server shutdown cleanly?
Thanks for any advice.
P.S. This thread has been moved from Servers > Planning to Servers > general - Hp Forum Moderator
Solved! Go to Solution.
02-16-2004 02:32 AM
The MSA500 take approximately 30 sec to 1 minute to initialize. If connecting to a ups, configure the ups to power on the MSA500 first and then the server 1 minute later. There sould be an option in the ups software.
If the server power up before or at the same time as the MSA500, storage may not be recognized by OS.
02-16-2004 02:44 AM
Thanks. What I'd not realised is that the Compaq UPS supports staged power-up on different outlets.
The small (1000VA) APC units I've used before don't support this.
04-03-2004 09:04 AM
You dont mention the OS vendor - if its Microsoft Windows, you could try this method.
In order to stagger the OS startup for each server, edit the TIMEOUT value in the C:\boot.ini file on each node.
I suggest setting node-A to 300 seconds (5 mins) and node-B to 420 seconds (7 mins).
This way the MSA 500 should have sufficient time to complete its start-up processes before node-A loads. By the time node-B starts, the cluster should be up and running on node-A.
I have used this many times in the past on NT4 EE & W2K A/S clusters and I dont see why it would not work on W2K3 clusters.
Hope this help you.