FHIR Bulk API resources limitation in response

[whitew...@gmail.com]

Hi All

I'm testing Bulk API against public HAPI server. BTW I have submitted the following kickoff-request:

{
  "resourceType": "Parameters",
  "parameter": [
    {
      "name": "_type",
      "valueString": "Observation"
    }
  ]
}

I got 10 files each one containing 1.000 Observations, then a total of 10.000 Observation.

To make a double check I submitted the followin non bulk query to the server:

http://hapi.fhir.org/baseR4/Observation?_total=accurate&_pretty=true
getting in the response a total of 4102153 Observation (read directly in the response after the activation of the _total=accurate rest parameter).

Are the bulk API limited somehow in total amount of returned resources?
I mean, which is the rational behind this difference?

Thank in advance for your help.

[James Agnew]

The public HAPI FHIR server is limited to a maximum of 10,000 results through any means of searching and exporting. We do this in order to prevent individual users from taking up so many server resources that it impedes other people from testing.

If you want to do large scale testing, I'd recommend deploying a private dedicated instance of HAPI FHIR and using Synthea to generate the amount of data you need.

Cheers,

James