diff --git a/Dockerfile.indexers b/Dockerfile.indexers index a3defb9..270d5d7 100644 --- a/Dockerfile.indexers +++ b/Dockerfile.indexers @@ -1,8 +1,6 @@ ARG BASE FROM $BASE -COPY swh-storage /app/swh-storage COPY swh-indexer /app/swh-indexer - -RUN pip install /app/swh-storage && pip install /app/swh-indexer +RUN pip install /app/swh-indexer diff --git a/kubernetes/72-indexers.yml b/kubernetes/72-indexers.yml new file mode 100644 index 0000000..ee1dea8 --- /dev/null +++ b/kubernetes/72-indexers.yml @@ -0,0 +1,140 @@ +--- +apiVersion: v1 +kind: ConfigMap +metadata: + name: indexers +data: + config.yml: | + storage: + cls: remote + url: http://storage:5002/ + objstorage: + cls: remote + url: http://objstorage:5003/ + indexer_storage: + cls: remote + url: http://idx-storage:5007/ + scheduler: + cls: remote + url: http://scheduler:5008/ + celery: + task_broker: amqp://guest:guest@amqp// + task_modules: + - swh.indexer.tasks + task_queues: + - swh.indexer.tasks.OriginMetadata + tools: + name: swh-metadata-detector + version: 0.0.2 + configuration: {} + + entrypoint-init.sh: | + #!/bin/bash + + set -e + + # echo Waiting for RabbitMQ to start + wait-for-it amqp:5672 -s --timeout=0 + + # echo Register task types in scheduler database + wait-for-it scheduler:5008 -s --timeout=0 + swh scheduler --url http://scheduler:5008 task-type register + + # Wait for services to be up + wait-for-it idx-storage:5007 -s --timeout=0 + wait-for-it storage:5002 -s --timeout=0 + wait-for-it objstorage:5003 -s --timeout=0 + + entrypoint.sh: | + #!/bin/bash + + set -e + + echo Starting the swh Celery worker + exec python -m celery \ + --app=swh.scheduler.celery_backend.config.app \ + worker \ + --pool=prefork --events \ + --concurrency=${CONCURRENCY} \ + --max-tasks-per-child=${MAX_TASKS_PER_CHILD} \ + -Ofair --loglevel=${LOGLEVEL} \ + --hostname "${HOSTNAME}" + +--- +apiVersion: apps/v1 +kind: Deployment +metadata: + name: indexers + labels: + app: indexers +spec: + replicas: 1 + selector: + matchLabels: + app: indexers + strategy: + type: RollingUpdate + rollingUpdate: + maxSurge: 1 + template: + metadata: + labels: + app: indexers + spec: + initContainers: + - name: indexers-init + image: swh/indexers:latest + imagePullPolicy: Always + command: + - /entrypoint.sh + volumeMounts: + - name: config + mountPath: /etc/softwareheritage/config.yml + subPath: config.yml + readOnly: true + - name: config + mountPath: /entrypoint.sh + subPath: entrypoint-init.sh + readOnly: true + containers: + - name: indexers + image: swh/indexers:latest + imagePullPolicy: Always + command: + - /entrypoint.sh + ports: + - containerPort: 5002 + resources: + requests: + memory: "256Mi" + cpu: "200m" + limits: + memory: "1024Mi" + cpu: "800m" + lifecycle: + preStop: + exec: + command: ["kill", "1"] + env: + - name: CONCURRENCY + value: "1" + - name: MAX_TASKS_PER_CHILD + value: "5" + - name: LOGLEVEL + value: "INFO" + - name: SWH_CONFIG_FILENAME + value: /etc/softwareheritage/config.yml + volumeMounts: + - name: config + mountPath: /etc/softwareheritage/config.yml + subPath: config.yml + readOnly: true + - name: config + mountPath: /entrypoint.sh + subPath: entrypoint.sh + readOnly: true + volumes: + - name: config + configMap: + name: indexers + defaultMode: 0777 diff --git a/skaffold.yaml b/skaffold.yaml index 79b99c9..9aa33b7 100644 --- a/skaffold.yaml +++ b/skaffold.yaml @@ -1,96 +1,97 @@ apiVersion: skaffold/v2beta13 kind: Config metadata: name: swh-environment build: local: useBuildkit: true concurrency: 2 artifacts: - image: swh/stack-base context: docker docker: dockerfile: Dockerfile - image: swh/objstorage docker: dockerfile: Dockerfile.objstorage requires: - image: swh/stack-base alias: BASE - image: swh/storage docker: dockerfile: Dockerfile.storage requires: - image: swh/stack-base alias: BASE - image: swh/scheduler docker: dockerfile: Dockerfile.scheduler requires: - image: swh/stack-base alias: BASE - image: swh/webapp docker: dockerfile: Dockerfile.webapp requires: - image: swh/stack-base alias: BASE - image: swh/loaders docker: dockerfile: Dockerfile.loaders requires: - image: swh/stack-base alias: BASE - image: swh/listers docker: dockerfile: Dockerfile.listers requires: - image: swh/stack-base alias: BASE - image: swh/grafana docker: dockerfile: Dockerfile.grafana - image: swh/search docker: dockerfile: Dockerfile.search requires: - image: swh/stack-base alias: BASE - image: swh/counters docker: dockerfile: Dockerfile.counters requires: - image: swh/stack-base alias: BASE - image: swh/indexers docker: dockerfile: Dockerfile.indexers requires: - image: swh/stack-base alias: BASE deploy: kubectl: manifests: - kubernetes/01-journal.yml - kubernetes/02-monitoring.yml - kubernetes/05-storage-db.yml - kubernetes/10-objstorage.yml - kubernetes/11-storage.yml - kubernetes/15-scheduler-db.yml - kubernetes/16-rabbitmq.yml - kubernetes/20-scheduler.yml - kubernetes/21-scheduler-runner.yml - kubernetes/22-scheduler-listener.yml - kubernetes/23-scheduler-journal-client.yml - kubernetes/29-web-db.yml - kubernetes/30-webapp.yml - kubernetes/40-loaders.yml - kubernetes/45-listers.yml - kubernetes/50-elasticsearch.yml - kubernetes/55-search.yml - kubernetes/56-search-journal-client.yml - kubernetes/60-redis.yml - kubernetes/61-counters.yml - kubernetes/62-counters-journal-client.yml - kubernetes/70-idx-storage-db.yml - kubernetes/71-idx-storage.yml + - kubernetes/72-indexers.yml