Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Have had enough consumer SSDs fail on me that I ended up building a NAS with mirrored enterprise ones...but 2nd hand ones. Figured between mirrored and enterprise that's an OK gamble.

Still to be seen how that works out in long run but so far so good.



For data storage, I just avoid SSDs outright. I only use them for games and my OS. I've seen too many SSDs fail without warning into a state where no data is recoverable, which is extremely rare for HDDs unless they're physically damaged.


SSDs are worth it to me because the restore and rebuild times are so much faster. Larger HDDs can take several days to rebuild a damaged array, and other drives have a higher risk of failure when they're being thrashed by IO and running hot. And if it does have subsequent drives fail during the rebuild, it takes even longer to restore from backup. I'm much happier to just run lots of SSDs in a configuration where they can be quickly and easily replaced.


I just don't have the patience for HDDs anymore. Mirrored arrays and backups are going to have to do on data loss.

That said I only have a couple of TBs...bit more and HDDs do become unavoidable


I'm using an HDD with SSD cache for /home all non stale will be cached by the SSD


What mechanism are you using to manage the cache ?


lvm


I’m worried about my drives that contain helium. So far so good, all show 100% helium level, but I wonder for how long.


Worst case they end up with normal atmosphere...which is what all the other drives are running anyway, no?

Can't say I've heard of people worrying about this angle before tbh


I understand normal atmosphere would destroy the heads pretty quickly.


You can't trust SSDs or HDDs, fundamentally they still have high failure rates regardless. Modern Filesystems with checksums and scrub cycles etc are going to be necessary for a long time yet.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: