Skip to content

Latest commit

 

History

History
82 lines (62 loc) · 2.73 KB

snowstorm_snomed.md

File metadata and controls

82 lines (62 loc) · 2.73 KB

Import SNOMED CT with NZ extensions to Snowstorm terminology server

Download SNOMED

Just to look: www.health.govt.nz A site where we can find all files we need (authorization required): https://mlds.ihtsdotools.org/#/dashboard

List of files to download:

  • SnomedCT_InternationalRF2_PRODUCTION_20200309T120000Z.zip
  • SnomedCT_ManagedServiceNZ_PRODUCTION_NZ1000210_20200401T000000Z.zip
  • SnomedCT_GPFP_PRODUCTION_20191030T120000Z.zip
  • SnomedCT_GPFPICPC2_PRODUCTION_20190412T120000Z.zip

The GPFP Refset is represented in both archives listed above.

Prepare archives

I’m not sure that it is necessary, but if separate imports don’t give a result, you can unpack all archives, collect all files from SNAPSHOT folders to one folder saving its directory structure and zip it.

Deploy Snowstorm

Clone a repository https://github.com/IHTSDO/snowstorm/

In docker-compose.yml change public port for snowstorm container to avoid conflicts:

snowstorm:
image: snomedinternational/snowstorm:latest

container_name: snowstorm

depends_on:
elasticsearch:
condition: service_healthy

entrypoint: java -Xms2g -Xmx4g -jar snowstorm.jar --elasticsearch.urls=http://es:9200

networks:
- elastic

ports:
- 8085:8080

Assure that your docker app has enough resources (in other case elastic search won’t start and you will see an error)

Snowstorm creates all ES indexes after starting.

Adjust Elasticsearch to avoid import errors

Possible import error: the number of terms [422770] used in the Terms Query request has exceeded the allowed maximum

To avoid the error while importing files execute this request against elastic:

PUT http://localhost:9200/_settings


{
    "index": {
        "max_terms_count": 1000000
    }
}

List existing indexes by executing http://localhost:9200/_cat/indices and check if the setting has been updated with http://localhost:9200/_settings

Import SNOMED archive

(according to instructions)

  1. Create an import job
{
  "branchPath": "MAIN",
  "createCodeSystemVersion": true,
  "type": "SNAPSHOT"
}

  1. Copy import id

  2. Upload the archive - The process can take more than one hour

  3. Assure that import job is COMPLETED

  4. Check the result using http://localhost:8085/fhir/ValueSet/$expand?url=http://snomed.info/sct?fhir_vs=ecl/%5E450970008%20AND%20%3C%3C64572001 as example