Hi managers,
I am trying to specify a small TruCluster 5.1 machine that would have full redundancy, but be as low-cost as possible. I am willing to withstand a short downtime when a failure occurs, but I want to eliminate as many SPOFs (Single Points of Failure) as feasible.
I am thinking of a couple of DS-10's, Memory Channel (I know 5.1a may eliminate this), and a couple BA-350 shelfs.
I think I can use LSM to mirror my data drives, /usr, and /var so there are no problems there.
The boot drive, quorum drive, and the root partition cannot be software mirrored to my knowledge.
Since these drives will tend to have pretty static data on them, I could back them up to spare drives with dd if I have similar spare drives.
My question is for each of the 3 disk types above what are the implications of a failure and what steps would be required to start using the dd'ed backup?
All 3 are in the hardware database, so it has to be updated, but if the drives are needed to boot the cluster, what do you do?
Secondly, if using BA-350 shelfs just won't get me to a reasonable solution, what is the least expensive HSZ solution (new and/or used would both be worth considering.)
I will summarize.
Thanks,
Greg Freemyer
Internet Engineer
Deployment and Integration Specialist
The Norcross Group
www.NorcrossGroup.com
Received on Thu May 17 2001 - 20:47:42 NZST