[HN Gopher] LVM Cache Surprises
___________________________________________________________________
LVM Cache Surprises
Author : JNRowe
Score : 34 points
Date : 2022-03-07 13:00 UTC (10 hours ago)
(HTM) web link (current.workingdirectory.net)
(TXT) w3m dump (current.workingdirectory.net)
| mfontani wrote:
| I've been debugging a sporadic high IO wait time very recently,
| and we ended up tuning vm.dirty_ratio and
| vm.dirty_background_ratio. I wonder whether those were already
| tuned from the very high % of ram default, or whether it could
| have any impact here.
| ColonelPhantom wrote:
| I use lvmcache for my desktop, which has an old and tiny SSD
| (most of which is given to Windows) but a large HDD. Performance
| can be unpredictable at times, but is generally quite good.
| antongribok wrote:
| I have about 50-100 PB of object storage workloads sitting on top
| of LVM cache for the past 5 years or so in production. It works
| extremely well.
|
| Before going into prod, we tested LVM cache, and bcache, and LVM
| cache worked better for our workloads. Also, the tooling is
| significantly better than bcache.
|
| One thing I love about LVM cache is if you are already running
| with LVM, you can add the caching layer on the fly without any
| downtime.
|
| Regarding the note at the end where the author sees poor
| performance. I recommend taking a very close look at the SSD
| model. I've seen enterprise SSDs that perform worse with sync IO
| than spinning disks. When we buy hardware, we tell our vendors
| the exact model of SSD they can use. If we haven't qualified the
| SSD, we don't put it in production.
| quadrature wrote:
| Could you expand a bit on what your use case is ?. Is this a
| cache in front of an s3 data warehouse or something like that
| ?.
___________________________________________________________________
(page generated 2022-03-07 23:01 UTC)