Ours is a mix, but the overwhelming majority are 'all tables'. There are a couple with 'primary table' that are either single-table entities or instances where we really do only want to see updates on the primary table. There are a couple with 'custom' that are built on views. The rest are 'all tables' as that is basically our default.
t 520.293.1488 ext. 7472
Glad to help. I don't believe the change tracking type is the source of the issues we've seen. Based on the evidence I've seen, it is more based on timing. For example, many of our missing items were updated within a particular date range, which suggests there was some hiccup that occurred during that time. Other missing items may have been similar, and doing a periodic 'full' refresh waylays those types of issues. It appears there was also an issue within the change tracking system itself as identified in KB 4611858:
This is available for version 10.0.20 and I'm working on installing it now but will probably leave the periodic full exports in place anyway just for safety.
Please let me know if I may provide additional information.
I believe we had to move away from incrementals and go to full pushes for the same reasons. This also introduces other problems but it's the best option we have until we plan our potential move to the data lake.
If you've found this thread useful, dive deeper into User Group community content by role