007revad
commented
1 year ago You could try enabling TRIM. But I suspect the process that deletes old snapshots and then recovers the now unused space would still cause your overheating issue.
I think getting some heatsinks is the best option.
You could set the fan speed to Full before doing any snapshot deletions then set it back to your preferred setting after - but that would become annoying to have to do regularly. If the NVMe drive(s) still overheat with the fan set to full then nothing you can change in the fan's config file will help - part from changing the shutdown temperature.
Your NVMe drives have a Operating temp of 0°C to 70°C so I would not increase the shutdown temperature about the current 70°C. The M.2 shutdown temp in DSM 7.2 on the DS1821+ is set to 71°C
There are a few settings that could help:
- Change the fan speeds so the fans increase their speed to full speed sooner.
- Change the 2nd and 3rd temperatures to lower values so the fans increase their at a lower temperature.
- Change the shutdown temperature to a higher temperature (not something I'd do unless I had checked the maximum temperature for my specific brand/model of NVMe drive(s). Your NVMe drives have a Operating temp of 0°C to 70°C so I would not increase the shutdown temperature about the current 70°C.
- The skip_check_temp="10" looks like it tells DSM how long to wait between temperature checks.
- The period="30" looks like how long DSM will run the fans at that speed before reverting back to checking the temperature.
With the M.2 drives located where they are I suspect it would take some time for them to benefit from an increase in fan speed.
The DS1821+ schemd.xml file is different to a Synology with an Intel CPU.
<adt_fan_config skip_check_temp="10" type="DUAL_MODE_LOW" hibernation_speed="UNKNOWN">
<alert_config threshold="2" period="30" alert_temp="68" shutdown_temp="71" name="m2"/>
<manual_config name="m2_temperature" threshold="2" period="30" temp="0" fan_speed="70"/>
<manual_config name="m2_temperature" threshold="2" period="30" temp="48" fan_speed="130"/>
<manual_config name="m2_temperature" threshold="2" period="30" temp="59" fan_speed="170"/>
<manual_config name="m2_temperature" threshold="2" period="30" temp="64" fan_speed="255"/>
aferende
Hi everyone, I open this thread because I had a high temperature problem on my NVME drives used in a RAID Pool on my NAS DS1821+. The disks are 2 WD Red SN700 2 TB (these are the technical specifications: https://www.westerndigital.com/it-it/products/internal-drives/wd-red-sn700-nvme-ssd#WDS200T1R0C) Especially the 2nd NVME (only the 2th strangely) reached 71 degrees and the NAS shut down for protection.
Searching the internet I found this post: https://www.reddit.com/r/synology/comments/12aci2n/disk_overheat_disk_m2_drive_1had_reached_70c/ where some NAS system files are indicated where it is possible to change the operating limit temperature.
I use these disks on a Pool dedicated to VMM and the temperature peak was reached when I went to delete some old snapshots. At that point the system started recovering the unoccupied space and made the temperature of the 2nd disk rise from 41 degrees to 71 degrees in a few seconds. (I haven't activated LTRIM, Do you suggest to activate it? Could improve this phase?)
Given that in the meantime I ordered 2 heat synks for NVME https://www.amazon.it/gp/product/B09NRKT3GW/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1 (hoping they improve the situation),
I would like to understand from you if you have encountered the same problem, if there is a way to insert in the script some option that modifies the temperature limit in NAS system config file (always if it is the case to do so) maybe as optional.
Thank you all