[tahoe-dev] Storage Node Proactive Share Relocation

[Nathan Eisenberg]

Hi all,

I got to wondering last night... say you have a grid that you're removing a storage node from for one reason or another.  In my mind, the current correct procedure would be to run a repair-walk through all the known DIRCAPs, to ensure everything is in a good state to start with, then stop the tahoe-LAFS storage node process, and run the repair walk again to regenerate the missing shares.

However, if you've got valid shares on that server, why not be able to use them to regenerate shares?  In other words, have the storage node reallocate its own shares before making them unavailable.

I realize this is probably not super important, but as I finish up the design of our production grid, I'm starting to think about server lifecycle, and how easy it will be to slide new servers in or pull old servers out without disrupting operations.  So any thoughts in general on that topic would certainly be appreciated.

Nathan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tahoe-lafs.org/pipermail/tahoe-dev/attachments/20110511/da48a420/attachment.html>