> 'Got that message from the console...
>
> - 2.0.32-1
> - Benjamin C.R. LaHaise's 2 inode patches + Nathan Bryant's fix on
> locking the wrong inode
> - 3000 filehandles
> - new 3c509.c from cesdis
I had to bang on the system pretty hard to do it, but I just managed to
trigger the same message. Same setup as above sans 3000 fd's and 3c509.
What triggered it: running several concurrent looping find /'s while
repeatedly copying /usr to /usr2 and deleting /usr2... and just for the
heck of it, rm -rf'ing /usr2 while it's being copied to... anyone got
other ideas on stressing the inode code?
Anyway I'll keep poking around to see what I can find out.
> The machine is still up and running for more than 4 hours now.
Ditto, just got one of the messages before I killed the tests,
everything's still running. __iget() seems intelligent enough about
recovering from this one...
Nathan Bryant
nathan@nbryan71.dorm.usm.maine.edu
nbryan71@mail.caps.maine.edu