Re: Enterprise replication
Posted in 1999
Madison Pruet — — source: Informix-list mailing list archive (1991-1998)
When a site has been down for a while and/or you are using time based replication, it is possible that quite a few messages might have gotten queued for a given site. When replication resumes for that site, it's possible that sending all of the data could flood the network. Therefor, there is a hold that occurs when a certain number of messages have been queued to a site. The that number drops to 5 outstanding messages, then the site will be resumed and more messages will be sent. Link, David A. wrote: > What is a holding state in the syscdrserver table?