FW: [Spread-users] spread exception

[Daniel Rall]

John Schultz <jschultz at d-fusion.net> writes:

> Just wanted to let the spread list know that we at D-Fusion have also
> run into this problem, and that the client failure gives no
> information about what went wrong.

Yeah.  :(

> I don't believe that it is our spread queue is getting overfull -- we
> take a lot of precautions to ensure this doesn't happen (recompiled
> spread to hold up to 10000 msgs per client, have a high priority
> thread always receiving, there aren't that many msgs being produced,
> etc.).
> 
> unfortunately, the problem was rare enough and our app was robust
> enough to handle such failures that we never dug into it to really
> figure out what was going wrong. I wouldn't be surprised if it was
> just a minor problem with the Spread Java client implementation.

We have a similar situation here.  I recompiled with a queue of 5000,
and have not hit the problem since.  I would like to know what's
actually going wrong wrong here, but find out hasn't yet become a
priority...
-- 

Daniel Rall <dlr at finemaltcoding.com>