thrroow Posted May 2, 2020 Share Posted May 2, 2020 When I transfer files to/from the array or use something like bittorrent to download files to the array (or really the cache SSD), between 3 and 5 of my threads will peg out at 100% usage in the unRAID dashboard. Checking top and htop shows that nothing is using that much CPU, however if I look at iotop or Netdata, I am showing very high IOWAIT times. It's like something is bogging down file transfers and it's crashing the entire system. After this goes on for a few seconds, all docker containers become unresponsive, VMs, and even unRAID dashboard. Attached is a picture with all containers and everything shut down except qbittorrent and 1 single torrent. CPU usage hasn't been nearly this high for such a task in the past. I had a similar problem in the last major version of unRAID, but upgrading seemed to have fixed it. It only now randomly re-appeared. Specs: 2x Xeon CPU 64gb memory 2x SSD cache drives in JBOD 9x HDD in array downloads set to save to cache first then get moved at a later date to array Quote Link to comment
bonienl Posted May 2, 2020 Share Posted May 2, 2020 (edited) 10 minutes ago, thrroow said: I am showing very high IOWAIT times High IO wait times are usually the result of poor cabling or seating. On the Main page you can click on the "disk-icon" for each device and it will show disk related events. Check for any errors. Edited May 2, 2020 by bonienl Quote Link to comment
JorgeB Posted May 2, 2020 Share Posted May 2, 2020 Also make sure docker image is on cache, there are reports of this happening if it's on the array, e.g.: Quote Link to comment
thrroow Posted May 3, 2020 Author Share Posted May 3, 2020 I believe it is this: Not sure why it would start just now, though Quote Link to comment
bonienl Posted May 3, 2020 Share Posted May 3, 2020 I have a cache pool of 4 SSD disks in RAID10 mode, and it is ultra fast. Quote Link to comment
Alexander Posted May 5, 2020 Share Posted May 5, 2020 I don't use a cache pool. When the parity check starts my Windows VM (on my array) practically freezes. I guess this is normal? Fix is to pause the parity check until I am finished with my VM. I guess another solution is to use cache or to put your VM "unprotected" on an unassigned device? Quote Link to comment
JorgeB Posted May 6, 2020 Share Posted May 6, 2020 8 hours ago, Alexander said: When the parity check starts my Windows VM (on my array) practically freezes. I guess this is normal? It's expect, you should have any vdisk outside the array for best performance, you can have it on a cache pool so it still remains protected. 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.