Hi Andy,

On Tue, Aug 24, 2004 at 12:09:30AM -0700, Andrew Hanushevsky wrote:
> You have to catch the situation at just the right time. I never did in my
> testing. The major difference is that in the real environment you have
> lots of clients asking for files and it's the combination of pending
> queries and servers comming and going that uncovers the bug. The code that
> introduced the bug wasn't there in previous releases.

  Ok. I did some basic checks and made a new version this morning:

  At least we are learning things on each of the full server restarts at 
SLAC. Clearly we need to return to the discussion of how to setup a large 
test system and just do it.


Peter Elmer     E-mail: [log in to unmask]      Phone: +41 (22) 767-4644
Address: CERN Division PPE, Bat. 32 2C-14, CH-1211 Geneva 23, Switzerland