UK Edition Loading Tips 🇬‍🇧

[kaicode]

Intro
The UK Edition of SNOMED CT requires some special handling. The edition is larger than most and some of the packages have an unusual format where there are multiple concept files within one zip file.

Tips
The UK Edition works very well in Snowstorm. Follow these tips for smooth loading of the RF2 package.

1. Use the Monolith Snapshot package
   • This is clearly labeled in TRUD "SNOMED CT UK Monolith Edition, RF2: Snapshot".
   • The download filename has the format uk_sct2mo_39.4.0_20250115000001Z.zip
   • Other distribution formats, for example with filename SnomedCT_UKEditionRF_PRODUCTION20250115.zip, will not load correctly.
2. Increase Snowstorm's max-terms count setting to 1 million
   • Use this setting: --elasticsearch.index.max.terms.count=1000000
   • This can be added at the end of the java startup command or included in a config file, see Configuration Guide.
3. Assign plenty of memory when loading the package
   • We know the monolith loads into Snowstorm well in Docker environment with 12g of memory where Elasticsearch has 6g and Snowstorm has 4g.
   • These are the recommended Docker Compose changes:

services:
  elasticsearch:
  ...
    environment:
    ...
      - "ES_JAVA_OPTS=-Xms6g -Xmx6g"
    ...
    mem_reservation: 6g

  snowstorm:
  ...
    entrypoint: java -Xms4g -Xmx4g --add-opens java.base/java.lang=ALL-UNNAMED --add-opens=java.base/java.util=ALL-UNNAMED -cp @/app/jib-classpath-file org.snomed.snowstorm.SnowstormApplication --elasticsearch.urls=http://es:9200 --elasticsearch.index.max.terms.count=1000000
  ...
    mem_reservation: 4g

Make sure Docker's Overall Memory Limit is set to 12g or above.

Happy Snowstorming! 🇬‍🇧