Re: [HACKERS] Performance while loading data and indexing

Neil Conway <neilc(at)samurai(dot)com> writes:
> I'm not really familiar with the reasoning behind ext2's reputation as
> recovering poorly from crashes; if we fsync a WAL record to disk
> before we lose power, can't we recover reliably, even with ext2?

Up to a point. We do assume that the filesystem won't lose checkpointed
(sync'd) writes to data files. To the extent that the filesystem is
vulnerable to corruption of its own metadata for a file (indirect blocks
or whatever ext2 uses), that's not a completely safe assumption.

We'd be happiest with a filesystem that journals its own metadata and
not the user data in the file(s). I dunno if there are any.

Hmm, maybe this is why Oracle likes doing their own filesystem on a raw
device...

regards, tom lane