avif-converter docker image

To build and use avif-converter locally in a Debian Docker container with all dependencies, change into app directory and run:

docker-compose up --build avif-converter

This should install all necessary dependencies, including manually building and installing all supported codecs as shared libraries. It will then start a web server attached to port 8080.

Note: This build process will take a while.

Caching with Google Cloud Platform

If you're using the Docker container with Cloud Run, you can optionally enable caching. This way you don't have to regenerate the same images every time from scratch. Cloud Storage buckets are used as a cache. Environment variable CACHE_TIMEOUT defines the object timeout in seconds. Zero means the object never expires. The default is 43200.

You can use the following shell script as a template to create a bucket with a lifecycle policy that cleans up expired objects every day.

#!/bin/sh
set -e

PROJECT_ID="my-project"
STORAGE_CLASS="STANDARD"
BUCKET_LOCATION="US-CENTRAL1"
BUCKET_NAME="my-cache-bucket"

# https://cloud.google.com/storage/docs/creating-buckets

gsutil mb -p ${PROJECT_ID} -c ${STORAGE_CLASS} -l ${BUCKET_LOCATION} -b on gs://${BUCKET_NAME}

LIFECYCLE_CONFIG_FILE=$(mktemp --suffix=.json)
cat > ${LIFECYCLE_CONFIG_FILE} <<EOF
{
   "rule":[
      {
         "action":{
            "type":"Delete"
         },
         "condition":{
            "daysSinceCustomTime":0
         }
      }
   ]
}
EOF
gsutil lifecycle set ${LIFECYCLE_CONFIG_FILE} gs://${BUCKET_NAME}
rm -- ${LIFECYCLE_CONFIG_FILE}

Then set GCP_BUCKET environment variable to point to the created bucket in your Cloud Run service.

gcloud run deploy my-cloud-run-service \
--image=gcr.io/my-project/my-container-image \
--allow-unauthenticated \
--concurrency=8 \
--cpu=4 \
--memory=8192Mi \
--set-env-vars=GCP_BUCKET=my-cache-bucket \
--platform=managed \
--region=us-central1 \
--project=my-project