Update HAPI SERVER in FOSPS

[joofio]

its now 6.6.0 and 6.10 at least is available. This version has a bug on the language search (creates problems for searching epis)

[gmej]

For ePI or IPS? there are two FHIR servers due to R4 and R5 resources. Anyway we would ike some help with this task. Our repo for FHIR is a fork we did some omnths ago from the original. What's the best practice to upgrade this server? Should we pull the changes made on the original fhir repo?

[joofio]

epi. You are using docker right?

[gmej]

Kuebernetes, which is the same at the end. We are deploying via the FHIR Helm Chart

[joofio]

hum, dont know for sure, but could you just update the docker? Is that feasbile?

[gmej]

Yes, we can, we will do it through this week or next one. We can take advantage and upgrade everything that is necessary. Is there any udpate on the IGs for GH? We are using the IPS IG v1.1.0, is there any update on that IG?

For the ePI server, are there any IGs we are missing to include?

[joofio]

IPS is stable. that is ok. With the updated HAPI, things will go smoother for search and query.

@jkiddo do you see any issue here that i may be forgetting? The server gets a clean install of content everyday.

[gmej]

Hmmm well, the server updates resources daily (the connector does it) but the deployment and configuration of the FHIR servers themselves is done via configuration that we have to apply and change, updates on that are not automatic.

With the udpates we get on this issue, we will update both FHIR servers and necessary and leave them ready forMVP3.

[joofio]

Seems like a good idea. But can we test in the dev environment before?

[jkiddo]

IPS is stable. that is ok. With the updated HAPI, things will go smoother for search and query.

@jkiddo do you see any issue here that i may be forgetting? The server gets a clean install of content everyday.

Nop - no issue as I see it

[joofio]

@gmej can we test on the dev then?

[gmej]

Yes of course, we should try this on dev first

[joofio]

great, let me know when you pick this up so i can help

[joofio]

will this be done this week?

[gmej]

Hi, sorry for delaying this that much.

Both epi and ips servers are upgraded to 6.10 only in dev.

There was an error with the following implementation guide when deploying the ePI server in application.yaml, it "couldn't be found":

    implementationguides:
      gh_0_1_0:
        url: https://build.fhir.org/ig/hl7-eu/gravitate-health/package.tgz
        name: hl7.eu.fhir.gh
        version: 0.1.0

Isn't it available anymore?

[joofio]

change the url to packageUrl and check again pls

[gmej]

Thanks, that was the solution. I am struggling with the connector on dev, I'll reach you again when resources are uploaded correctly to dev, as we found a bug in the connector.

[gmej]

Could you check this version upgrade solves the bug related to language search? Contents are already present in the development environment.

[joofio]

I was talking to @jkiddo and I think some configs are needed still. Will get back ASAP

[jkiddo]

HAPI FHIR JPA Starter 7.0.1 is what you want.

[gmej]

I see v7 was released just yesterday! https://github.com/hapifhir/hapi-fhir-jpaserver-starter/releases/tag/image%2Fv7.0.0

Then should we update to 7 now? Is this necessary for now?

[joofio]

lets wait for 7.0.1

[joofio]

@jkiddo I believe the upgrade can be done now and should be solved right?

[jkiddo]

Yes - use 7.0.2

[joofio]

wait not yet

[joofio]

dont have the option ig_runtime_upload_enabled=true. With that false and the option "language_search_parameter_enabled: true" should run ok

[gmej]

So can we udpate it then or should we wait for something? I don't know about those two options, where should we write that?

[joofio]

you can update. the options are for application.yml

[gmej]

both fhir servers? epi and IPS? regarding this, isn't there a way to use only one FHIR server? Are there any plans on doing this?

[jkiddo]

Wait for 7.0.3 to be released

[joofio]

only epi is required. but you can do both if it is easier. IPS is only FHIR R4

[joofio]

Wait for 7.0.3 to be released

why? We dont need the IG run time here

[jkiddo]

If you cant wait 10 minutes then ok

[joofio]

ah ok! Thats great then! Thanks. letts wait for it then

[gmej]

I will upgrade to 7.0.3 tomorrow or monday. If anything should be taken into account I'll be watching this issue.

Also if there are any plans of new versions soon, we can wait and avoid upgrading many times.

[joofio]

no. the language part is quite important. But for now i dont see any issue in the foreseeable future at least that would requite upgrade

[gmej]

@jkiddo I synced the fork, built correctly, but got this log when FHIR was starting:

  .   ____          _            __ _ _
 /\\ / ___'_ __ _ _(_)_ __  __ _ \ \ \ \
( ( )\___ | '_ | '_| | '_ \/ _` | \ \ \ \
 \\/  ___)| |_)| | | | | || (_| |  ) ) ) )
  '  |____| .__|_| |_|_| |_\__, | / / / /
 =========|_|==============|___/=/_/_/_/
 :: Spring Boot ::                (v3.2.0)

2024-03-12 11:51:37.636 [background-preinit] INFO  o.h.validator.internal.util.Version [Version.java:21] HV000001: Hibernate Validator 8.0.0.Final
2024-03-12 11:51:37.711 [main] INFO  ca.uhn.fhir.jpa.starter.Application [StartupInfoLogger.java:50] Starting Application using Java 17.0.10 with PID 1 (/app/main.war started by nonroot in /app)
2024-03-12 11:51:37.713 [main] INFO  ca.uhn.fhir.jpa.starter.Application [SpringApplication.java:653] No active profile set, falling back to 1 default profile: "default"
2024-03-12 11:51:40.350 [main] INFO  o.s.d.r.c.RepositoryConfigurationDelegate [RepositoryConfigurationDelegate.java:139] Bootstrapping Spring Data JPA repositories in DEFAULT mode.
2024-03-12 11:51:40.809 [main] INFO  o.s.d.r.c.RepositoryConfigurationDelegate [RepositoryConfigurationDelegate.java:208] Finished Spring Data repository scanning in 446 ms. Found 53 JPA repository interfaces.
2024-03-12 11:51:41.068 [main] WARN  o.s.b.w.s.c.AnnotationConfigServletWebServerApplicationContext [AbstractApplicationContext.java:632] Exception encountered during context initialization - cancelling refresh attempt: org.springframework.beans.factory.support.BeanDefinitionOverrideException: Invalid bean definition with name 'subscriptionQueryValidator' defined in ca.uhn.fhir.jpa.subscription.submit.config.SubscriptionSubmitterConfig: Cannot register bean definition [Root bean: class [null]; scope=; abstract=false; lazyInit=null; autowireMode=3; dependencyCheck=0; autowireCandidate=true; primary=false; factoryBeanName=ca.uhn.fhir.jpa.subscription.submit.config.SubscriptionSubmitterConfig; factoryMethodName=subscriptionQueryValidator; initMethodNames=null; destroyMethodNames=[(inferred)]; defined in ca.uhn.fhir.jpa.subscription.submit.config.SubscriptionSubmitterConfig] for bean 'subscriptionQueryValidator' since there is already [Root bean: class [null]; scope=; abstract=false; lazyInit=true; autowireMode=3; dependencyCheck=0; autowireCandidate=true; primary=false; factoryBeanName=ca.uhn.fhir.jpa.topic.SubscriptionTopicConfig; factoryMethodName=subscriptionQueryValidator; initMethodNames=null; destroyMethodNames=[(inferred)]; defined in class path resource [ca/uhn/fhir/jpa/topic/SubscriptionTopicConfig.class]] bound.
2024-03-12 11:51:41.086 [main] INFO  o.s.b.a.l.ConditionEvaluationReportLogger [ConditionEvaluationReportLogger.java:82] 

Error starting ApplicationContext. To display the condition evaluation report re-run your application with 'debug' enabled.
2024-03-12 11:51:41.120 [main] ERROR o.s.b.d.LoggingFailureAnalysisReporter [LoggingFailureAnalysisReporter.java:40] 

***************************
APPLICATION FAILED TO START
***************************

Description:

The bean 'subscriptionQueryValidator', defined in ca.uhn.fhir.jpa.subscription.submit.config.SubscriptionSubmitterConfig, could not be registered. A bean with that name has already been defined in class path resource [ca/uhn/fhir/jpa/topic/SubscriptionTopicConfig.class] and overriding is disabled.

Action:

Consider renaming one of the beans or enabling overriding by setting spring.main.allow-bean-definition-overriding=true
2024-03-12T12:51:41.129356798+01:00

[jkiddo]

@gmej see https://github.com/hapifhir/hapi-fhir-jpaserver-starter/issues/655

[gmej]

Thanks for the fast response. ePI server in development is already upgraded to v7.0.2

[joofio]

@gmej the parameter is "language_search_parameter_enabled: true" is it enabled?

[gmej]

Done

[joofio]

did you restart?

[gmej]

Yes, isn't this feature available? Have you checked you are trying in development?

[joofio]

yes

HTTP/1.1 400 Bad Request
x-powered-by: HAPI FHIR 7.0.2 REST Server (FHIR Server; FHIR 5.0.0/R5)
x-request-id: 432bbb72-fa79-9b1a-b23e-1316573bae65
content-type: application/fhir+json;charset=UTF-8
date: Mon, 18 Mar 2024 11:09:17 GMT
x-envoy-upstream-service-time: 66
server: istio-envoy
connection: close
transfer-encoding: chunked

{
  "resourceType": "OperationOutcome",
  "issue": [
    {
      "severity": "error",
      "code": "processing",
      "diagnostics": "HAPI-1223: Unknown search parameter \"_language\" for resource type \"Composition\". Valid search parameters for this search are: [_id, _lastUpdated, _text, attester, author, category, date, encounter, entry, event-code, event-reference, identifier, patient, period, related, section, section-code-text, section-text, status, subject, title, type, url, version]"
    }
  ]
}

how did you put the parameter? the full name is hapi.fhir.language_search_parameter_enabled=true

[gmej]

My bad. Now it should be working, I am able to search with "_language", even though 0 results are returned

[joofio]

it now works. there is the need to reindex for these things to work:

reindex

POST https://gravitate-health.lst.tfo.upm.es/epi/api/fhir/$reindex

Good Job!