On Mon, Apr 16, 2007 at 12:40:37PM -0400, Nathan Kennedy wrote: > > I think the good time to do this is as soon as possible; what we really > need to be afraid of is a second failure on our other drive before we get > the first one replaced.
Supposing that that will not happen, we still need to choose an appropriate time for downtime, and let users know in advance. > Make sure Interserver has non HCoop email and phone contacts for during > the replacement, in case the server does go down. Give them the model > number and disk size of the drives from dmesg, so they can swap a disc > with ideally the same model. Yes, I am planning to do so. > I'm not sure how well software raid would > perform if we had two drives of different geometries or spin rates even if > they were the same size. I think there will be no problem with different geometries since RAID works on partition level... So in case our partitions are not exactly matching with the new disk, we only have to make them a few megabytes bigger on the new disk to be sure, and re-add to existing array. Cya -doc _______________________________________________ HCoop-Discuss mailing list [email protected] http://hcoop.net/cgi-bin/mailman/listinfo/hcoop-discuss
