Re:We saved millions in SSD costs by upgrading our filesystem
by abrookewood on Tuesday November 09, 2021 @09:08PM
(#29170304)
That's a great point. Potentially they could implement regular replacement of nodes to obtain the benefits of de-fragmented pools.
Re:We saved millions in SSD costs by upgrading our filesystem
by throwdbaaway on Wednesday November 10, 2021 @03:35AM
(#29172429)
Indeed. It looks like they went from ZFS 0.8.x (lz4) to 2.x (zstandard) in one go. That would be too yolo for my taste, but the guys were under time pressure due to the rather sudden performance degradation. Nodes replacement would probably buy some time, to allow the upgrade to be done in 2 phases.
Re:We saved millions in SSD costs by upgrading our filesystem
by Ford__Prefect on Tuesday November 09, 2021 @10:18PM
(#29170675)
For read frag on ssd shouldn matter, and for write since zfs is tree struct the allocator would create few more nodes and links but no big deal ?
Re:We saved millions in SSD costs by upgrading our filesystem
by ericbarrett on Tuesday November 09, 2021 @11:05PM
(#29170958)
Link is 404 for me.
Re:We saved millions in SSD costs by upgrading our filesystem
by phaer on Wednesday November 10, 2021 @02:41AM
(#29172100)
The link seems to be truncated above: https://utcc.utoronto.ca/~cks/space/blog/solaris/ZFSGangBloc...
This is probably not a fair comparison.
On the existing lz4 machines, the zfs pools are already badly fragmented, making it difficult to find regular blocks: https://utcc.utoronto.ca/~cks/space/blog/solaris/ZFSGangBloc...
On the new zstd machines, the zfs pools are still pristine, after being restored from backup.
If we want to isolate the effects of migrating from lz4 to zstd, we also need some new lz4 machines for comparison.