Igor Raits wrote on Fri, Jun 05, 2020:
zramctl shows ALGORITHM
NAME ALGORITHM DISKSIZE DATA COMPR TOTAL STREAMS MOUNTPOINT
/dev/zram0 lzo-rle 11.7G 4K 74B 12K 8 [SWAP]
So it is lzo-rle by default, but it should be possible to override this
algorithm. There is an RFE for this already at zram-generator github.
There is zstd support in our kernels now so that would be good to
compare, I'm seeing impressive results but as this is workload-dependent
it is hard to test from an objective point of view, but this is rather
impressive:
# zramctl
NAME ALGORITHM DISKSIZE DATA COMPR TOTAL STREAMS MOUNTPOINT
/dev/zram0 zstd 1G 753.4M 150.6M 158.8M 8 [SWAP]
(enabled 15 or so days ago manually; according to smem I have bits of
qemu, firefox, libreoffice, xorg, urxvt and most of lightdm,
sd-pam, rpc.mountd.. in swap so it is quite varied)
Either way +1 from me, I've just installed the generator and manually
set the limit to 4G/0.5 as suggested in the change; I think such a
config file should be added to the f32 rpm so people can try more
easily to convince themselves with the actual config that would be
installed.
Cheers,
--
Dominique