Rework database triggers to use FlowRun.delete_from_counts
#3702
Conversation
Codecov Report
@@ Coverage Diff @@
## main #3702 +/- ##
=========================================
Coverage 100.00% 100.00%
=========================================
Files 488 489 +1
Lines 26167 26168 +1
=========================================
+ Hits 26167 26168 +1
Continue to review full report at Codecov.
|
| responded=bool([e for e in run["events"] if e["type"] == "msg_received"]), | ||
| ) | ||
| else: | ||
| FlowRun.objects.filter(id=run_obj.id).update(**db_state) |
There was a problem hiding this comment.
updating this to match more closely how things work in mailroom
| ---------------------------------------------------------------------- | ||
| -- Increments or decrements our start counts for each exit type | ||
| ---------------------------------------------------------------------- | ||
| CREATE OR REPLACE FUNCTION temba_update_flowstartcount() RETURNS TRIGGER AS $$ |
There was a problem hiding this comment.
Moved the INSERT part into temba_flowrun_insert. We don't UPDATE start_id on flow runs except to trim starts.. at which point is more efficient to just delete the counts. We also don't need to update counts if runs are being deleted, and that might help make archiver deletes a little more efficient.
So from now on, only result category counts will be decremented when a run row is deleted. Path, exit type and start counts remain unchanged. Less trying to rewind history.