Mike ShoeboxOriginal post
Web Developer
1.) Is this still this is a valid strategy for having parallel processing of async function executions? I saw some support threads on the appwrite site which were unclear to me.
2.) If I have 2 workers processing function executions in parallel, are the instances isolated from each other?
e.g. if I have one exceution running which downloads files locally to operate on them, I do not want the 2nd parallel execution to overwrite these files or do any operations on the files from the other function execution. I want to be 100% sure these executions have their own namespace on disk (I'm not referring to persistent storage btw -- just parallel execution using additional worker-functions).
appwrite-worker-functions-1:
image: appwrite/appwrite:1.5.7
entrypoint: worker-functions
# <<: *x-logging
container_name: appwrite-worker-functions-1
restart: unless-stopped
# networks:
# - appwrite
depends_on:
- appwrite-redis
- appwrite-mariadb
- openruntimes-executor
environment:
- _APP_ENV
- ... more envs
appwrite-worker-functions-2:
image: appwrite/appwrite:1.5.7
entrypoint: worker-functions
# <<: *x-logging
container_name: appwrite-worker-functions-2
restart: unless-stopped
# networks:
# - appwrite
depends_on:
- appwrite-redis
- appwrite-mariadb
- openruntimes-executor
environment:
- _APP_ENV
- ... more envs
...
Summary
- Developer wants to know if adding duplicate worker-functions in docker-compose.yml is a valid strategy for parallel processing of async function executions.
- Developer seeks clarification on whether instances of parallel executions are isolated from each other to avoid file conflicts.
- Make sure each worker-function has its own namespace on disk to prevent overwriting or interference with files from other executions.