On Wed, Jan 15, 2003 at 12:34:18PM -0800, nate wrote: > he did mention that doing a fsck will take 16 hours
I guess he did. > that and if his data is so critical to remain available I think his > organization needs to come up with a better system configuration > with more high availability configurations in mind Sounds like a real mess. The major disruption that they are likely going to have might turn into the occasion to upgrade to a configuration that has a hope of being as reliable as they need. I am impressed with the constructive suggestion of doing very selective surgery manually to possibly fix things and keep them running long enough to transition to a more suited system. Another possibility: Quickly (!) craft a redundant and reliable system and somehow run the two together. The new system would sit in front, talking to users, saving new data in reliable disks. The old system would sit behind--in a read only mode so corruption won't get worse--supplying old data to the new system. Once all the data has migrated to the new system, breath a sigh of relief and pull the plug on the old beast! What a nightmare, -kb, the Kent who was too harsh in his earlier "hold your nose and go down for 16-hours" reaction. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe https://listman.redhat.com/mailman/listinfo/redhat-list
