salcio

@macavelly have you ever resolved that issue ? I'm having very similar behaviour.

That's sound awesome. It will defiantly help to split the caches per use case. Do you have any thought about the current setup ? Like why do i see drop in performance (small but still) between software (btrfs) RAID10 and hardware (P840) RAID 10?

With the current setup (4SSD through P840 RAID controller) there is only one Drive visible to unRaid. So no "Cache Pools" from unraid point of view. Are there any changes in 6.9 that affect cache.

Hi, I've got a HP Gen9 D380 server with P840 smart array controller. At the beginning I've connected all the drives through on board ports so unraid can see them directly. I've set up a cache pool with 4SDD in btrfs RAID10 configuration. It worked ok (benchamark screenshots attached [SSD Metrics]). Then I though I should get better cache performance if I use hardware RAID10 though this P840 controller. And that what I've done - all 4 SSD are in RAID10 configuration on smart array. In unraid drive is formatted with BTRFS and set as cache drive. But the performance did not improve (decreased actually - screenshots attached []). Benchmarks were run from the VM with img file hosted on cache drive. So few questions: 1. Is having a cache drive done through P840 a good idea ? (Pros cons) ? 2. Is BTRFS a good choice for cache in this configuration - maybe XFS would be more performant ? 3. In the banchmarks I'm not sure about the Acc. time - should that be something in range of 0.030 - 0.100 ?

Cool. Thank you very much.

Do I need to assign them in any particular order ? or does it not matter ?

So I have something similar I thing. I installed new raid controller and set it to hba mode (for now). After restart my cache drives are not assigned (but visible in unassigned devices) - data and parity disks were assigned correctly. Can I just assign disks to cache again. It was a cache pool with 4 drives.

@johnnie.black @John_M have you ever find out why there is a difference between writes to /mnt/diskX and /mnt/cache ? I'm seeing significant differences in speeds - using your tests: cache 10240000+0 records in 10240000+0 records out 10485760000 bytes (10 GB, 9.8 GiB) copied, 51.2255 s, 205 MB/s user ^C1225556+0 records in 1225556+0 records out 1254969344 bytes (1.3 GB, 1.2 GiB) copied, 177.72 s, 7.1 MB/s Writing to the same share. Share set up as using cache only. I'm using unRAID Version: 6.8.3.