Help get this topic noticed by sharing it on Twitter, Facebook, or email.

Seemingly random occurrences of missing timestamp columns

This is a bit of an odd one, but I'm hoping I'm not completely crazy in my theory.

We have a scheduled SQL job that refreshes a set of critical tables locally every 10 mins. Many internal systems depend on these tables. Every once in a while, we will see a spike in processes blocking each other and, lately, we've been able to trace this to DBAmp replicating a table that shouldn't be replicated.

When checking the logs, we see the message "Table does not contain a timestamp column needed to refresh. Using SF_Replicate to create table." Looking at the SF_Refresh proc, this appears to be triggered if no timestamp column can be found for the table being refreshed, which makes sense. What doesn't make sense is that all of the tables in our critical list have SystemModStamp.

Is it possible that some intermittent connection issues could lead to this behavior?

I've added some additional text to the log messages to spit out whatever value is in the @timestamp_col_name var, in hopes of getting some more data, but this doesn't happen very often. When it does, though, it causes all sorts of trouble by locking up the DB during critical business hours.
1 person has
this question