Yeah, I’ve disabled markdown-oxide for the moment, so I’ll see what my uptimes look like for a bit.
I honestly can’t imagine how a userspace program could cause this behavior, though. There’s no memory pressure, and there are 16 cores in this CPU, fer chrissake. Even trying to peg the CPU, I didn’t notice the md-oxide correlation until I started watching top; the temps weren’t going up, performance wasn’t impacted.
I thought for sure it was a memory (hardware) issue, but I’ve run several memtests and they come back clean. No odd kernel module crashes in the logs; no indication anything is wrong until - poof. Reboot.
Yeah, I’ve disabled markdown-oxide for the moment, so I’ll see what my uptimes look like for a bit.
I honestly can’t imagine how a userspace program could cause this behavior, though. There’s no memory pressure, and there are 16 cores in this CPU, fer chrissake. Even trying to peg the CPU, I didn’t notice the md-oxide correlation until I started watching top; the temps weren’t going up, performance wasn’t impacted.
I thought for sure it was a memory (hardware) issue, but I’ve run several memtests and they come back clean. No odd kernel module crashes in the logs; no indication anything is wrong until - poof. Reboot.