A simple drop_caches fixes the massive latency spikes on allocation (until the page cache grows to use all free RAM again, of course). This is obviously broken. Ugh.
Show this thread
-
Disks dying is a normal failure mode. A hardware glitch corrupting RAID6 stripes during disk replacement necessitating a roll back to the old disk and thus a patchwork of old and new disk state isn't. ZFS works great until it doesn't and then you can't fix it.
"ZFS doesn't need fsck" is what they want you to think. Then something happens they don't plan for and you're screwed. Go look around for ZFS data loss stories, they exist.
-
-
I'm fully aware of the data loss stories. Most of the losses stems from improperly configured storage pools.
-
"improperly configured" = "you did something we didn't expect or plan for". I'd rather have a filesystem that at least attempts to provide recovery tools for when you make a "mistake", thanks.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.