That hints at one of my problems with the idea of
using it as
concurrent service. I'll elaborate:
1. Maintaining up-to-date realtime mirroring would be a HUGE
SUCKING SOUND on bandwidth. This, it seems to me, is not an option.
Does anyone know the actually volume of updates in (MB/GB) per day. With
some smart replication, you could simply pass a diff between sites to avoid
the complete transmission of articles; and pass a CRC for the final article
from the originating server; so you can verify that the article has been
replicated properly. If the replication fails then you pass the complete
article across.
//Eden