Stream: bulk data
Topic: Stream processing
Jeff Graham (May 02 2018 at 14:59):
Hello everyone, just learned about this discussion group from Josh and Dan at the HIMSS Dev4Health conference. Nice to see work being done on bulk fhir data movement. Our data science team does a lot of clinical data movement (non-fhir atm) and we've had great results routing data over kafka. Would like to see how we could contribute to the streaming design.
Josh Mandel (May 02 2018 at 15:02):
Thanks Jeff! When you think about streaming, are there's aspects you'd want to standardize (as in, via standardize messaging protocol) vs sharing best practice and maybe your for working with specific tools like Kafka?
Josh Mandel (May 02 2018 at 15:02):
Put another way: what are the problems you'd like to see HL7 tackle?
Jeff Graham (May 02 2018 at 15:11):
Probably both. First things that come to mind are standardizing on a source publish approach, message protocol, schema registry for handling changes to fhir, sink connectors to things like analytical databases et al.
Jeff Graham (May 02 2018 at 15:12):
Then best practices on handling changes to data, capacity planning, stream transforms, etc.
Grahame Grieve (May 06 2018 at 03:57):
I'd like to see this happen too - but recall that the EHR vendors, at least, had strong resistence to this kind of approach
Last updated: Apr 12 2022 at 19:14 UTC
