I'm just getting a cluster of Coda clients and serves up to speed.
Three servers, about 10 clients, about 10 users.  Everybody's home
directories served over Coda, with a (probably insecure) strategy of
automatically giving a token to every UID every day.

Two of the clients are served over wireless Ethernet with a bandwidth
of about 1.1 Mbps (practically; that's for ftp connections within the
local network).  It's slower than Real Ethernet but not *that* slow.

For some reason, running a make on a wireless-served client on a big
shared (coda-served) source tree just completely hoses the client.
I'd expect it to be stressful.. lots of files deleted and created,
lots of files changing content.  In any case, the client often gets in
a state where "ls" and "ls -l" show completely different sets of
files, and one or two files get marked as inconsistent and need to be
fixed with repair. Sometimes repair works, sometimes it doesn't.
Also, I get a lot of this "Cannot create regular file: Connection
timed out."  This is a pain in the butt.  What's going on?

In general, is there any way to say "Force synchronization of all
volumes right now, and I mean it?"  Seems like reintegration messages
pop up VERY slowly when changes are happening fairly rapidly.  I'd
like clients on multiple machines to have a somewhat-uniform view of a
shared filesystem, or at least some way to speed up synchronization
when it's important.

coda clients and servers 5.2.7, Linux 2.2.10, kernel-venus module
5.2.3.   

Thanks,
Bill Gribble