You are not logged in.
Pages: 1
I have just created a RAID1, using two Seagate 2TB disks. I have changed the speed_limit_min and speed_limit_max to have them sync faster (which I believe is reset after a reboot), other than that, the RAID1 config is standard. The disk is also encrypted using dm-crypt (as ar all my disks).
When I copied files to this new RAID, it occasionally stalls for a second or two, and then continued copying. And just recently I took a backup of it, using rsync, to an external disk, and the same thing happened. It stalls occasionally for a second or two.
This does NOT happen between any other disks, only when interacting with the RAID.
Does anyone have any idea on how I can fix this?
Offline
Still having this problem and haven't found any fix to it. Anyone who has experienced the same?
Offline
Is there anything in dmesg when it stalls? smartctl -a looks fine and both disks pass a smartctl -t long self-test?
Encryption can cause stalls, if the CPU is sufficiently busy. Although it should not be noticable with AES on a AES-NI system...
Offline
What are you using for the RAID? When I used on-board fake-RAID controller from Intel I frequently had this problem. The solution, for me, was to go pure software RAID.
Also, it may be that the array is stalling due to some kind of snag disk-side. Consumer grade HDDs will try to read from a drive until it absolutely fails to do so, instead of telling the controller to read from another disks/use parity. This, too, can cause stalls, as well as drive drop outs.
Offline
Pages: 1