Interview Questions
1. In Merge replication, for the conflict resolver: I have a row guid and a timestamp column on each article in the publication. It was my hope that by having the timestamp, I could avoid the need to manually reconcile the conflictsbetween publisher and subscriber. However, I see that the conflicts are still there and still require manual intervention to eliminate. Why?
Merge Replication does NOT require a timestamp column (that’s for transactional replication). Merge Replication only
requires a ROWGUID (which must be the ROWGUIDCOL for the article). Second, the conflict resolver didn’t work the way I
expected it to: the conflict resolver works by using a default rule to resolve a conflict, but it still records the fact that a conflict
occurred. The record of the conflict (and the winning and losing records) are stored (by default) at the distributor. Manual
intervention is required (opening the conflict viewer) to view and eliminate the conflicts. But if you want to be alerted to the
existence of a conflict, you must write a separate monitoring tool.
Submit an interview question
Submitted questions and answers are subject to review and editing, and may or may not be selected for posting, at the sole discretion of w3Sniff.
Get Started
Comments
Leave a Comment