deployed FhirConverter with deployhl72fhir.bash script

when checking the logs I see this error:

2022-03-15T17:44:22.472Z INFO - Pulling image: healthplatformregistry.azurecr.io/fhirconverter:v2.1.1
2022-03-15T17:44:23.825Z ERROR - DockerApiException: Docker API responded with status code=NotFound, response={"message":"manifest for healthplatformregistry.azurecr.io/fhirconverter:v2.1.1 not found: manifest unknown: manifest tagged by "v2.1.1" is not found"}

2022-03-15T17:44:23.826Z ERROR - Pull image threw Exception: Input string was not in a correct format.
2022-03-15T17:44:23.827Z INFO - Pulling image from Docker hub: healthplatformregistry.azurecr.io/fhirconverter:v2.1.1
2022-03-15T17:44:25.024Z ERROR - DockerApiException: Docker API responded with status code=NotFound, response={"message":"manifest for healthplatformregistry.azurecr.io/fhirconverter:v2.1.1 not found: manifest unknown: manifest tagged by "v2.1.1" is not found"}

2022-03-15T17:44:25.025Z WARN - Image pull failed. Defaulting to local copy if present.
2022-03-15T17:44:25.032Z ERROR - Image pull failed: Verify docker image configuration and credentials (if using private repository)
2022-03-15T17:44:25.033Z INFO - Stopping site hl72fhirv4hl7conv3489 because it failed during startup.
/home/LogFiles>

previous version worked fine.

Hi,

Tried deploying HL7Conversion:

• Deploying FHIRProxy creates a KeyVault
• Deploying Ingest creates a KeyVault
• When deploying Convert (deployhl72fhir.bash), script is asking for one KeyVault name instead of the 2 above, causing the script to fail.

Hi all,

Set up a fhir server and a fhir proxy and followed instructions in readme. When I try to authorize the FHIRServerProxy-1 API Connection. Is the fhirdeployhl72.bash suppose to add replyurl config to the app registration of the logic app?

I'm not sure if I am confusing 2 APIs but as I was following the instructions to set up the Participant Filter Post-Processor I kept getting a 404 when trying to link my user to a fhir resource.

The instructions said to call the API endpoint following the path
https://<your fhir proxy url>/api/manage/link/<ResourceName>/<ResourceID>?name=<user principal name>.

however looking through the code it looks like the correct path is
https://<your fhir proxy url>/api/manage/link/<ResourceName>/<ResourceID>/<user principal name>

Are the readme instructions incorrect?

Hi,

If the Client Secret has a special char / or + or…Logic App in FHIRExportwithAnonymization fails, even though the same Client Secret works for others like Azure API for FHIR, HL7Conversion...

Trying to retrieve Function App Host Key when Function App is not fully created yet.
Adding a sleep 10 helped. But opening and saving in notepad, is causing pipe failed.

Here is a comment

Still testing automations

Still testing more team automation

Still testing flows

Still testing

I dont see a way to have my MLLP gateway send traffic to a store and use the Liquid templates? Is there a way to do that?
Thank you,

We ran across an issue where the Azure API for FHIR returns a continuation token that is larger than 4kb. The maxQueryStringLength parameter of Azure Functions is set to 4kb, so anything larger than that is truncated.

Does the FHIRProxy handle this scenario?

See related issue: microsoft/fhir-server#1567

I would like to confirm a particular use case processing details.
The reference code is in hl7ingest.cs.

Use Case:
Input HL7 message Type MDM T02 with an OBX segment.

Given that : "The OBX segment can also contain encapsulated data, e.g., a CDA document or a DICOM image"
We have seen in practice that the OBX segment contains large size( lets say for example 10 MB) base64 encoded docs or images.

Question: Does the hl7ingest.cs code preserve the strict order for all messages?

To be more precise since the processing function will first save the message to a blob storage and then queue , it would appear that is possible for a new(smaller in size) HL7 to be received, processed and queued before the large MDM message loaded in storage and queued.

I wonder how this use case can be handled in the ingest component or is something that the HL7 conversion component should address.

Last Step in Logic Apps, Import Bundle to FHIR Server is not working.

I found this issue while working with deployhl72fhir.bash and confirmed that the same issue exists in at least one more .bash script. The start of the scripts has set -euo pipefail, however this is not honored in the sub-shells. This results in confusing behavior, like scripts saying they succeeded when they actually failed.

I don't have time for a PR now (apologies), but wanted to submit this for tracking. Probably the easiest solution is to implement what is suggested here and require Bash >= 4.4 since Bash 4+ is required already.

Example of what the user sees when this issue arises:

After running the deployhl72fhir.bash script everything is set up.

After that FHIRServerProxy-1 API Connection is in Error state ( This connection is not authenticated )

When attempting to Authorize I am directed to a login form for my credentials and getting this error message after logging in:

AADSTS90008: The user or administrator has not consented to use the application with ID '3648ca46-8d1e-4d62-8cde-27c00e70b03b'(https://sfp-proxy18447.azurewebsites.net). This happened because application is misconfigured: it must require access to Microsoft Graph by specifying at least 'Sign in and read user profile' permission.

In the documentation it is stated :

%%%%%
Goto the resource group that contains the deployed HL7toFHIR workflow
Find the FHIRServerProxy-1 API Connection and click on it
Click on Edit API Connection
Enter a Display name of FHIR Server Proxy
Click Authorize
Enter the Credentials for the user account, group or service principal from step #4
Click Save
%%%%%

3648ca46-8d1e-4d62-8cde-27c00e70b03b was created by the scripts and the values added to the key vault.

How can be this issue resolved? Where should I "Enter the Credentials for the user account, group or service principal from step #4"?

Hi :) When we have looked for the Fhir proxy solution it was hard to find. It could be good to have it in the main readme. We are also willing to make a contribution on that subject.

DependsOn on /Batch_Config needs to be changed to "[concat(variables('factoryId'), '/linkedServices/AzureKeyVault')]" or is failing.

Hello, I'm trying to us configure the services endpoint using the documentation, but the when i run the docker command file (using windows as container) the command didn't work. when change to Linux container, the command run, but when i try to run this command => docker run -it --rm -p 8080:8080 myfhirvalidator-svc ... an error is shown and the image never run....

Also, when running the validator downloaded from here http://hl7.org/fhir/validator/ some of the profiles
i've being validating say "This is not a resource" or it does validate and throw errors like "Property not valid on
us-core...". Any help would be highly appretiated.

There are any consideration that i have to know.... in order to run this properly?

Regards !

Hi,

We are running into issues finalizing the SMART Proxy setup for external users within the Azure API for FHIR. We have created roles and are at the stage of Linking Users in the Participant Roles to FHIR Resources but are getting a 404 error. Can you please explain where one would grant the FHIR Administrative Role to complete the set-up.

health-architectures/FHIR/FHIRProxy at master · microsoft/health-architectures · GitHub

Kind Regards,
Moses Chege
Solution Architect , CHC