nvme timeouts

[myitcv]

Using the tech preview of VMWare Fusion for Silicon, I had been seeing a number of disk related hangs on my machine. This generally coincided with my doing something relatively disk intensive, like a find, recursive grep etc. Indeed I could reliably reproduce this issue after a fresh boot via a git grep in a large repository.

I happened to glance at /var/log/syslog and saw a number of lines like:

    nvme nvme0: I/O 9 QID 7 timeout, aborting

Switching the root hard disk bus type to be SATA instead of NVMe fixed the issue 100%. 

This all occurred whilst using the 5.14.21-051421-generic kernel

So I'm reporting this here in case a) it's a bug and b) others run into the same issue.

Thanks

[myitcv]

Confirming this problem remains with Ubuntu 22.04 and kernel v5.16.2.

[myitcv]

For anyone following along, conversation also happening at https://twitter.com/_myitcv/status/1486023011060916225

[myitcv]

FYI this problem appears to be fixed with VMWare Fusion 22H2 preview (20191287)

[myitcv]

Cancel that, this is still an issue with VMWare Fusion 22H2 preview (20191287)