Hacker News Reader

Most SSD's can't be trusted to maintain proper data ordering in the case of a sudden power off.

That makes checksums and journals of only marginal usefulness.

I wish some review website would have a robot plug and unplug the power cable in a test rig for a few weeks and rate which SSD manufacturers are robust to this stuff.

View on HN

I'd say it makes checksums even more important so that you know whether something got corrupted immediately and not after a year (or whatever) has gone by and you actually need it.

The problem is that if the SSD suffers a power failure and reverts a 1 megabyte block of metadata to the way it was yesterday, the filesystem won't see that as corruption - since all the checksums will match.

Yet all the pointers in that metadata will point to data which no longer exists, and your filesystem will be destroyed.

So... metadata checksums? I mean if enough data/metadata gets corrupted you're SOL either way.

For example, in ZFS the metadata is checksummed such that pointers to the data carry the checksum for that data.

Reply on news.ycombinator.com