Hi all,
We've always had a good number of failed merges each week due to the good ole "activity happened on the kept account" error. Sometimes I can see that something physically happened and so the error is legitimate (not that I always agree with it failing the merge because of that... but that's a different story). I also noticed a while ago that the LP_CUSTOMER_RANK was causing a lot of unwanted activity. Since we don't currently use ranking I disabled it which kind of helped.
But I find myself looking into this issue again as we've still been having a crazy number of merges failing each week and most of them are consistently the same people. This is what I've found and I'm looking for some advice on what other people have done.
The majority of the failures seem to be happening because of our nightly Manage Constituency updates. It's funny because most of these people have had the constituency for several months and the last_update_dt in TX_CONST_CUST is from several months ago. But their last_update_dt in T_CUSTOMER is for the exact time of the last constituency update run. How have other people gotten around this? Should I stop using the Manage Constituencies report and write up LPs to apply constituencies that don't modify the account unnecessarily?
This also leads me to wonder how I'll handle the ranking procedure and merges when we start using ranking in the future...
Thanks for any advice!
Hi Beth,
The merge process looks at t_customer.last_activity_dt to determine if there has been activity since the last merge. It compares that against t_potential_dups.last_update_dt.
Is there anything you are doing in your nightly jobs that is updating t_customer.last_activity_dt?
Thanks,David
We had a lot of failures in the past for a very similar reason (that is nightly jobs stepping on the merge job).
Hi Gawain,
This sounds like it is no longer an issue... So I'm curious as to how you resolved it! Did you change how your nightly jobs work or was there some sort of customization to the merge procedure?
Thanks! Beth