Method for replication tracing

摘要

In a network in which local copies of a shared document are maintained by multiple servers, performance data is gathered and stored during document synchronization via replication. The performance data includes a delay indicator which is associated with the document such that each copy of the document includes a record of which servers and hops were traversed and what delay was experienced at each of those servers and hops. The delays may be calculated by associating a time stamp with the document upon receipt at a server, and subsequently comparing time of receipt at a logically adjacent server with the time stamp. The performance data can be used to locate problems.