If #1 was solved by using the raid approach, what happens if one of the disks containing one of my table spaces crashes.
if you are using raid, your tablespaces are on raid volumes comprised of 2 or more drives, any one of those drives may fail, and the full data is still available. if you insist on keeping your tablespaces on seperate volumes, you'll need enough disks to have a raid for each tablespace, thats a minimum of 2 times the tablespaces in drives. putting all the drives in one large raid10 (striped mirrors) and putting all your data on that same volume will tend to distribute the IO as well as anything.
now, if you're asking what happens when a complete raid volume fails, you could consider that a system failure, and fail the whole mess over to a streaming standby server.... that would catch more failure cases and be much simpler and more robust.
-- john r pierce N 37, W 122 santa cruz ca mid-left coast -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
