sir_lou

After a lot of trial and error I believe I found the culprit. My Home Assistant has sensors that update in high frequencies. I did some tuning, exclusions and now I am waiting for the results. For future reference I used the Addon(App) DbStats to help me track the entities/sensors that caused a lot of writes. I ended doing some modifications to my configuration.yaml and now it looks like this: # Configure a default setup of Home Assistant (frontend, api, etc) default_config: # Optimize NVMe writes and DB degradation recorder: commit_interval: 60 purge_keep_days: 7 exclude: domains: - sun - weather - zone entity_globs: - sensor.time* - camera.brand_* - sensor.*rx* - sensor.*tx* - sensor.device_power_station_* - sensor.*_feels_like entities: - sensor.router_cpu_utilization - sensor.router_memory_utilization - sensor.homeassistant_rx - sensor.homeassistant_tx - sensor.iot_device_rx - sensor.iot_device_tx - sensor.rx - sensor.tx - media_player.living_room_media_player - sensor.server_cpu_usage - sensor.server_ram_usage - sensor.server_ram_used - sensor.server_cpu_power - sensor.server_cpu_temperature - sensor.server_network_inbound - sensor.server_network_outbound - sensor.server_disk_disk1_usage - sensor.server_disk_disk2_usage - sensor.server_disk_disk3_usage - sensor.server_disk_nvme_usage - sensor.server_disk_nvme2_usage

I will try tonight, check the difference in writes and post my reply tomorrow.

Hello everyone, I am trying to find out what is causing 50GB per day writes on my NVMe disks. Here is my setup. I have Unraid 7.3.1 installed on a AOOSTAR WTR PRO Ryzen. I have two pools, one with 4 mechanical drives and a ZFS cache pool with 2 nvme drives. (Compression is also on) In the cache pool I only host Home Assistant and a FreePBX. FreePBX is barely utilized, 5-10 calls per month. There are other VMs also but everything is Powered off for days. Home assistant is working for around 30 devices. Nothing else resides in the pool, no docker.img, no logs, no shares, nothing, I moved everything out. Here are the numbers that SMART reports for a single disk June 30 22:30 -Data units read1,682,618 [861 GB] -Data units written4,777,679 [2.44 TB] July 1 10:30 - Data units read 1,702,711 [871 GB] - Data units written 4,832,070 [2.47 TB] I believe this is excessive and I am looking for the culprit. Does anyone have an idea? File activity plugin does not show anything for the nvmes.