rushing-mallet7748 Posted October 12, 2023 Share Posted October 12, 2023 Hi, i have the problem that my unraid server becomes really unresponsive from time to time. The Processor view often shows all cores to be used by nearly 100%: But when i look into htop everything seems to be fine: I've ran Proxmox before on this server where i had no performance issues at all. Does anyone know what could cause this? Is it just caused by parity calculations? If thats the case, why can't i see them in htop or somewhere else? My system currently: OptiPlex 7070 Intel(R) Core(TM) i5-9500 32GB Ram DDR4 WD Blue SN570 2TB (Parity) Crucial BX500 2 TB (Disk) 18 Docker containers 2 Users Thank you Quote Link to comment
rushing-mallet7748 Posted October 12, 2023 Author Share Posted October 12, 2023 tower-syslog-20231012-1857.zip tower-diagnostics-20231012-2057.zip Quote Link to comment
JorgeB Posted October 13, 2023 Share Posted October 13, 2023 Does it return to normal if you stop docker? Quote Link to comment
rushing-mallet7748 Posted October 13, 2023 Author Share Posted October 13, 2023 Seems like it does. I've deactivated docker now in Settings -> Docker. After that the usage pretty much goes to zero. What could cause this? Quote Link to comment
JorgeB Posted October 13, 2023 Share Posted October 13, 2023 Start the containers one by one to see if you can find the culprit. Quote Link to comment
rushing-mallet7748 Posted October 13, 2023 Author Share Posted October 13, 2023 Okay, all containers are now stopped and i've started one of the smallest containers i have (homarr) The Usage immediately jumps up to 50%. Just for a short time but multiply this by 18 and it makes sense that there is such a high usage. But why is it like this? Under Proxmox with all containers enabled there were no performance issues at all. Even when i open the webpage of homarr in my browser i can see that there is a significant increase in usage in this graph. Quote Link to comment
JorgeB Posted October 13, 2023 Share Posted October 13, 2023 If you are using /mnt/user paths try with disk shares, or use an exclusive share. Quote Link to comment
rushing-mallet7748 Posted October 13, 2023 Author Share Posted October 13, 2023 I think i found the problem. I moved the docker directory to a dedicated pool drive. Now this is not part of the array anymore and it runs fine Quote Link to comment
Solution JorgeB Posted October 13, 2023 Solution Share Posted October 13, 2023 Yeah, docker in the array will be super slow, docker on a pool can still be faster with disk or exclusive shares if you find it slow. Quote Link to comment
rushing-mallet7748 Posted October 13, 2023 Author Share Posted October 13, 2023 What is the reason for this slowdown? Quote Link to comment
JonathanM Posted October 14, 2023 Share Posted October 14, 2023 4 hours ago, rushing-mallet7748 said: What is the reason for this slowdown? Realtime parity calculation and updates for all writes. Quote Link to comment
rushing-mallet7748 Posted October 14, 2023 Author Share Posted October 14, 2023 And why can't i see this in for example htop? Quote Link to comment
JonathanM Posted October 14, 2023 Share Posted October 14, 2023 htop doesn't show iowait normally, the dashboard includes it. 1 Quote Link to comment
rushing-mallet7748 Posted October 14, 2023 Author Share Posted October 14, 2023 ok, thanks for the clarification 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.