I'm getting reproducible kernel lockups with bees 0.6.3 after hours of successful dedup. The file system being deduped is not my root and not being accessed by anything but bees. My filesystem is newly created on DM thin-provisioning (LVM thin volume) on MD raid10, the thin pool metadata is on a different device (SSD MD raid1).
uname -a shows:
Linux onyx-z68 5.4.85-1-lts #1 SMP Mon, 21 Dec 2020 19:28:35 +0000 x86_64 GNU/Linux
It seems like the kernel soft-locks at first, but quickly gets worse and worse until it never returns control. I can't make sysrq's, ping it, or set caps lock on. I'm thinking it must be stuck in an interrupt handler or critical code section. I'm trying to get debugging info for you, but since I can't trigger a kdump and nothing seems to be sent to the kernel ring buffer (or at least to my dmesg tail over ssh) before I lose control I'm struggling.
I'm getting reproducible kernel lockups with bees 0.6.3 after hours of successful dedup. The file system being deduped is not my root and not being accessed by anything but bees. My filesystem is newly created on DM thin-provisioning (LVM thin volume) on MD raid10, the thin pool metadata is on a different device (SSD MD raid1).
uname -ashows:It seems like the kernel soft-locks at first, but quickly gets worse and worse until it never returns control. I can't make sysrq's, ping it, or set caps lock on. I'm thinking it must be stuck in an interrupt handler or critical code section. I'm trying to get debugging info for you, but since I can't trigger a kdump and nothing seems to be sent to the kernel ring buffer (or at least to my dmesg tail over ssh) before I lose control I'm struggling.