[[6.8.3] docker image huge amount of unnecessary writes on cache]

Someone should really make a PSA for this. I purchased a brand new ssd in January, 163 TBW on it now. Like others have mentioned earlier, iotop shows loop2 is constantly writing to the disk. Using df -kh shows /var/lib/docker is mounted on loop2.

 

I have a 5 disk btrfs encrypted array with a non-encrypted btrfs cache disk (Samsung EVO 860 1TB). Running several dockers, mainly for web hosting (traefik, cloudflare, organizr, etc.) and data storage (ms SQL Server, influxdb). No VMs.

Below is a slightly modified/simplified version of a script to calculate drive TBW and health %. Source here

#!/bin/bash

### replace sdg below with label of drive you want TBW calculated for  ###
device=/dev/sdg

sudo smartctl -A $device |awk '
$0 ~ /Power_On_Hours/ { poh=$10; printf "%s / %d hours / %d days / %.2f years\n",  $2, $10, $10 / 24, $10 / 24 / 365.25 }
$0 ~ /Total_LBAs_Written/ {
   lbas=$10;
   bytes=$10 * 512;
   mb= bytes / 1024^2;
   gb= bytes / 1024^3;
   tb= bytes / 1024^4;
   #printf "%s / %s  / %d mb / %.1f gb / %.3f tb\n", $2, $10, mb, gb, tb
     printf "%s / %.2f gb / %.2f tb\n", $2, gb, tb
   printf "mean writes per hour:  / %.3f gb / %.3f tb",  gb/poh, tb/poh
}
$0 ~ /Wear_Leveling_Count/ { printf "%s / %d (%% health)\n", $2, int($4) }
' |
   sed -e 's:/:@:' |
   sed -e "s\$^\$$device @ \$" |
   column -ts@