On Wed, 18 Sep 2002, Greg Troxel wrote:

>   Yes, I have seen this sometimes.  The repair just wedges or returns an
>   error and after that the client seems to be a quivering wreck
>   requiring an init to get it right.
>
> OK - so I am not the only one seeing this bug.

I suppose this is the problem I've seen too. Jan commented on it a couple
of times. His comment was that it is a known issue but hard to fix
properly.

Let us see the coming 6 release?

My approach is to make massive updates in "cfs strong" mode only.
It works fine "most of the time". I can still trigger disconnection by
arranging a heavy server load, but that server *is* slow - a P100 with
just 64M RAM (RVM is three times bigger) and old slow SCSI disks...

Disconnection does not necessary cause problems but with massive updates
it creates a possibility of trouble, as you say :-/

[A hint: "while sleep 1;do cfs cs;done" in background helps to avoid
disconnections in "extremely important" cases. YMMV :-]

Regards,
--
Ivan