Reason for change
Several ingest scripts have been running in parallel when large deliveries have been received.
When this happens the average time of execution goes from a few minutes to over one/two hours.
The other side effect seems to be that the whole database grinds to a halt, slowing down everything else.
A fix was added to lock the tables preventing ingestions to run in parallel.
They will be now serialized and, given that they run only a few minutes, the impact on data processing should be minimal.
A better fix could be added in the future if needed.
Test Procedure
This version has been tested on the DEV server.
Rollback procedure
Version 1.0.1 can be easily put back in place.
CCB Jira
Details
Jira Issues | ||
---|---|---|
|