Stream: implementers
Topic: Real Time Replication
Ovidiu Girlan (Jan 07 2019 at 18:22):
As a relatively new member, I am not sure if this topic has been discussed. If covered by another stream/topic I would really appreciate a pointer. I would try to provide a quick snapshot of the problem we try to tackle. Currently the Registered Persons Database (RPDB) and Corporate Provider DB (CPDB) are stored on a mainframe platform. There are various external systems that consume data (analytics mostly). Current solution is to replicate data (near-real-time) into a new DB which is used for querying. The idea is to build a FHIR capable ODS, which will expose data to external systems. The challenge is to keep this FHIR repo in sync with source DBs (mainframe RPDB and CPDB). Considering that source data uses specific schema, there is a need for a translation/conversion before updating fhir ODS. I wonder if there are any bets practices, lesson learnt during various projects facing similar problems.
Regards
ovidiu
Grahame Grieve (Jan 08 2019 at 19:53):
it's hard to generalise. Best practices - stable identifiers. Good logging. Stable identifiers.
Last updated: Apr 12 2022 at 19:14 UTC
