Importing Data

In order to use Blaze for queries, you have to import your data. Although you can use the full functionality of the FHIR RESTful API available under http://localhost:8080/fhir to create resources, the easiest way is to download blazectl to upload bundles.

First you should test connectivity by counting already available resources in Blaze which should be zero:

blazectl --server http://localhost:8080/fhir count-resources

which should return:

Count all resources on http://localhost:8080/fhir ...

-------------------------------------
total                             : 0

After that, you need a FHIR bundle to upload. You can generate one by downloading bbmri-fhir-gen and run:

mkdir fhir-test-data
bbmri-fhir-gen fhir-test-data

That will generate two files under fhir-test-data:

-rw-r--r--  1 akiel  staff    31K Nov  8 10:16 biobank.json
-rw-r--r--  1 akiel  staff   757K Nov  8 10:16 transaction-0.json

After you have the test data, you can upload it with:

blazectl --server http://localhost:8080/fhir upload fhir-test-data

which will output:

Starting Upload to http://localhost:8080/fhir ...
Uploads          [total, concurrency]     2, 2
Success          [ratio]                  100.00 %
Duration         [total]                  3s
Requ. Latencies  [mean, 50, 95, 99, max]  2.414s, 3.313s, 3.313s, 3.313s 3.313s
Proc. Latencies  [mean, 50, 95, 99, max]  2.412s, 3.31s, 3.31s, 3.31s 3.31s
Bytes In         [total, mean]            123.60 KiB, 61.80 KiB
Bytes Out        [total, mean]            788.30 KiB, 394.15 KiB
Status Codes     [code:count]             200:2

counting the FHIR resources again:

blazectl --server http://localhost:8080/fhir count-resources

should return:

Count all resources on http://localhost:8080/fhir ...

Condition                         : 100
Observation                       : 431
Organization                      :  11
Patient                           : 100
Specimen                          : 100
---------------------------------------
total                             : 742

Disabling Referential Integrity Checks

If your primary date source ensures referential integrity already, and you can't import the resources into Blaze in the right order or bunldes so that the referential integrity will be maintained per transaction, you can disable the referential integrity checks in Blaze by setting the env var ENFORCE_REFERENTIAL_INTEGRITY to false.

The checks can be enabled/disabled at each start of Blaze. You can check whether the checks are enabled by executing:

curl -s http://localhost:8080/fhir/metadata | jq -r 'isempty(.rest[].resource[].referencePolicy[] | select(. == "enforced")) | not'

That command return true or false depending on whether the referential integrity checks are enabled or not.

Please be aware that Blaze doesn't check the referential integrity of already imported resource if you enable the checks after you have already imported resources. By disabling the referential integrity checks at some time, you will have a database that can contain stale references.