On testing/bookworm, since booting on a 6-versioned linux-image, I have
noticed frequent hang ups of the nfs server, rendering it mostly
unusable. This is accompanied with Kernel Oops in the dmesg.
This sounds similar to previous bugs #1014793 and #1020548, both RESOLVED by
later patches.
How easy can you trigger and reproduce the issue? If you can easily
reach that situation, can you try to bisect the issue? Easiest would
be to first pin point between Debian revisions, and later further
bisect in upstream stable series.
Do you have the possibility to do that?