Suppose I have a storage space with two-way mirror based on 2 physical USB drives a and b.

Then I disconnect b and connect it to another computer.

Then I use—write some data—to both degraded storage spaces on both computers.

Then I connect drive b back together with drive a on the same computer.

  • Will “Repairing” cycle happen after this procedure?
  • Will data from drive a take preference after “Repairing” cycle?
  • Will it make difference if before reconnecting them together, I disconnect both a and b?
  • Will data from last disconnected drive take preference after “Repairing” cycle?
  • Will data from last written drive take preference after “Repairing” cycle?
  • Will data be merged by taking last modified 256Mb slabs from both drives?
  • Will data be merged by taking last modified file system cluster from both drives?

Same question about the same procedure with hardware RAIDs.

  • 1
    With respect of RAID system, look up STONITH, and how problems with 2 drives are handled - I'm betting that the RAID will not rebuild until you manually tell 1 drive to overwrite the other - exactly how you do that will depend on the hardware. – davidgo Mar 3 '15 at 7:25
  • And here is what STONITH means: “STONITH is a technique for NodeFencing, where the errant node which might have run amok with cluster resources is simply shot in the head. Normally, when an HA system declares a node as dead, it is merely speculating that it is dead. STONITH takes that speculation and makes it reality. "Make it so, Number One".” – Giacomo1968 Mar 3 '15 at 8:06
  • Your post contains too many separate questions. This is not a good fit for our Q&A format, so please create separate posts for each distinct issue you have. See also: One post with multiple questions or multiple posts? – bwDraco Mar 3 '15 at 17:20

These are some interesting questions, mixed with some odd ones... But this seems to be getting outside the intended use of Storage Spaces. This is really something you should consider just experimenting with if you can.

RE "same thing for hardware RAID": for almost all hardware RAID, the "current configuration" is stored on the controller and the drives. Once 1 drive is removed from a mirror, the config on the controller and remaining drive are updated to reflect that the other disk is a "failed" member of the mirror. If you take that drive elsewhere and try to make changes on it, then bring it back and reconnect it, its config will have been modified by the alternate controller and the original will mark it as "foreign". With a foreign configuration, you are forced to choose 1: clear the foreign config to keep the original config, or "import" to take the modified drive's config as the new good one, which would invalidate the original config on the other drive (and clear it).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.