Hi All
Currently we have some 3-peer replication setup(SQL 2012). Everything works fine, all tracers can be inserted and we have a total latency of less than 10 seconds. However, once in a while we can see Performance Critical in the replication monitor (up to 10 minutes latency). When we view the details we can't see any pending commands. This will fix itself randomly after some time (usually 30 mint or 2 hours). During all of this time everything just worked fine.
I am curious as to how the latency is calculated. Has anyone seen the same behavior and if it is save to ignore this.
Cheers