lukedyue Posted March 25 Share Posted March 25 Hello, I am using 3 SSDs for my Array with ZFS as filesystem, I always notice there are a lot of write requests to my first disk, I used to use `iotop` to find which process is writing to it, but it turns out to be `shfs`, that's hard to identify the real source? Recently I've noticed that the data units written would soon exceed the SSD's warranty (229/1500 TBW), and the written data is larger than disk2 and disk3 even it stores less data, so it's kinda weired. Is something wrong with ZFS or other setups? Could you please give some suggestions on this? Thanks a lot. Quote Link to comment
itimpi Posted March 25 Share Posted March 25 You should post your system's diagnostics zip file in your next post in this thread to get more informed feedback. It is always a good idea to post this if your question might involve us seeing how you have things set up or to look at recent logs. Quote Link to comment
lukedyue Posted March 25 Author Share Posted March 25 8 hours ago, itimpi said: You should post your system's diagnostics zip file in your next post in this thread to get more informed feedback. It is always a good idea to post this if your question might involve us seeing how you have things set up or to look at recent logs. Sorry, I've attached the diagnostics, could you please take a look at it? Thanks a lot! luke-nas-diagnostics-20240326-0054.zip Quote Link to comment
itimpi Posted March 25 Share Posted March 25 Disk1 appears to be where you have your 'system' and 'appdata' shares located and these will be used any time you have the docker subsystem or docker containers running which will be why you get lots of writes to that disk. Quote Link to comment
caplam Posted March 26 Share Posted March 26 back in unraid 6.8.3 i had a similar issue. Docker was running from a btrfs pool. Writes on the pool killed my ssds in 8 months. I managed to lower writes by - using directory for docker instead of image - use a single xfs pool - add --no-healthcheck as extra parameters for most of my containers - and use a ram disk for dockers logs. This one is great thanks to @mgutt: link to the topic Besides of this i also bought longer life ssds. Depending of my activity i have around 50 running containers. Quote Link to comment
lukedyue Posted March 31 Author Share Posted March 31 On 3/26/2024 at 4:34 AM, itimpi said: Disk1 appears to be where you have your 'system' and 'appdata' shares located and these will be used any time you have the docker subsystem or docker containers running which will be why you get lots of writes to that disk. Thanks! So I guess I need to use a durable SSD for disk1 or migrate docker files to another disk Quote Link to comment
lukedyue Posted March 31 Author Share Posted March 31 On 3/26/2024 at 5:12 PM, caplam said: back in unraid 6.8.3 i had a similar issue. Docker was running from a btrfs pool. Writes on the pool killed my ssds in 8 months. I managed to lower writes by - using directory for docker instead of image - use a single xfs pool - add --no-healthcheck as extra parameters for most of my containers - and use a ram disk for dockers logs. This one is great thanks to @mgutt: link to the topic Besides of this i also bought longer life ssds. Depending of my activity i have around 50 running containers. Wow, thanks for providing so much detail! I will try these! Thanks a lot! Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.