On Tue, Sep 14, 1999 at 01:49:03PM -0400, squeegy_at_squeegy.org wrote:
>  I got the following errors while filling my ner replicated volume:
... 
> 13:26:55 Reintegrate: s.src, 100/2293 records, result = SUCCESS
> 13:27:00 Fatal Signal (11); pid 11821 becoming a zombie...
> 13:27:00 You may use gdb to attach to 11821
> 
> wiggles:/export # ps ax | grep 11821
> 11821  a2 S    5:15 venus -init
> 
> the transfer of files has frozen.  I built my root volume in a

Well, not many programs can successfully handle a coredump. And venus
is doing exactly as it says. It became a zombie (living dead), waiting
for someone to attach a debugger, or to drive a stake through its heart
(killall -9 venus).

> partition that is 1.6GB.  The files I am copying total 232MB.  I
> should have plenty of room on my volume right?  The backup server's
> root partition is even larger.

Yes but Coda decided that the server was being sluggish, and switched to
write-disconnected operations. So it started logging operations, and
trickle reintegrate them back to the server. You possibly do not have a
232MB local cache.

If you use `cfs strong' before starting the copying, you tell venus to
ignore observed network conditions, and in 99% of the cases this should
avoid Coda to switch to (write-)disconnected operation.

> 13:44:42 Probably another Venus is running! open(/dev/cfs0) failed
> (1), exiting

There is, the `dead' venus is still holding on to the kernel. And the
mountpoint is probably still not unmounted. (umount /coda).

Jan