NVMe cache and crashed volume

[grimkroton]

Hi there,

I was hoping, that with the scripts, I would be able to reboot my DS423+ with 2x Kingston SNV3S1000G in a volume as cache without crashing the volume. Unfortunately, the volume is still crashing on reboots after running the script with version v3.6.119 and triggering it on reboot via the scheduled job.

Should the script fix this issue or did I have a wrong expectation?

[007revad]

With a DS423+ a cache on 3rd party NVMe drives should work and survive a DSM update without the script. So you don't need the script.

There's something else going on. For a NVMe cache (especially a read/write cache) I would be using NAS or Enterprise NVMe drives. They have a much higher lifespan, TBW limit and PLP (power loss protection).

I really hate NVMe read/write caches (especially with pinned meta data!) because your whole HDD volume is at risk if something happens to the NVMe drives.

You could shutdown the DS423+, remove the NVMe drives, blow out any dust in the M.2 slots, insert the NVMe drives and make sure they are seated correctly.

I would also check the SMART values for the NVMe drives with https://github.com/007revad/Synology_SMART_info

[grimkroton]

Maybe my last comment was a little confusing. The NVMes handle a DSM update but not a reboot. Everytime I reboot the NAS, the volume crashes. I need a poweroff/poweron cycle to have a consistent volume.