| 1 |
On Friday 26 May 2017 10:36:40 Kent Fredric wrote: |
| 2 |
> On Wed, 24 May 2017 15:45:45 +0300 |
| 3 |
> |
| 4 |
> Andrew Savchenko <bircoph@g.o> wrote: |
| 5 |
> > - smaller CPU overhead: not every i/o is being compressed, e.g. if |
| 6 |
> > there is sill enough RAM available it is used without compression |
| 7 |
> > overhead as usual, but if memory is not enough, swapped out pages |
| 8 |
> > are being compressed instead of swapping out to disk; |
| 9 |
> |
| 10 |
> I found the opposite problem somehow. CPU started becomming frequently |
| 11 |
> pegged in zswap for no obvious reason, while the underlying IO that zswap |
| 12 |
> was doing was only measurable in kb/s , far, far, far below the noise |
| 13 |
> thresholds and by no means a strain on even my crappy spinning rust based |
| 14 |
> swap. |
| 15 |
> |
| 16 |
> And to add to that, zswap introduced general protection faults and kernel |
| 17 |
> panics. |
| 18 |
> |
| 19 |
> So nah, I'm glad I turned that off, it was a huge mistake. |
| 20 |
|
| 21 |
Did you also have zbud enabled at the time? |
| 22 |
|
| 23 |
-- |
| 24 |
Regards, |
| 25 |
Mick |
Archives