Adam D. Moss
Adam D. Moss
I think it's an improvement, but... at the risk of reopening a maybe-never-closed can of worms: have you (or has anyone) successfully tried the advice offered on the FAQ? I've...
> My understanding is that the expectation is it works fine unless you're swapping due to low memory, since you might need to allocate to swap, and that goes Poorly....
I have a small worry, which is that this may be intentional (or 67% intentional...) to catch memory corruption in-place. But probably not.
A manual `zpool reopen` takes the pool to the `UNAVAIL` state and stops those spinning threads, which is progress! Unfortunately `zpool reopen` doesn't actually *return*, and outstanding zpool/dataset operations (zpool...
Finally got a flamegraph of the spinning - 
(as an aside, the comments here talk about a >1MB recordsize a few times - I thought 1MB was the upper limit?)
Apologies since I lost track; did this land in a state compatible with existing Linux copy_file_range/reflink support (i.e. cp --reflink) or is a separate flag/syscall/api needed to utilize this? Thanks!
> > did this land in a state compatible with existing Linux reflink support (i.e. cp --reflink) or is a separate flag/syscall/api needed to utilize this? Thanks! > > @adamdmoss...
It happens with fifo/wt, fifo/wb, and lru/wb
Removing the assert appears to make enhanceio work okay, but it doesn't make me feel confident. :)