Graphql db sync snapshot faild with "cannot drop objects owned by role ...."

Developers Misc Dev Talk
1

I tried to spin up a graphql environment from snapshot, but I failed. From genesis no problem, but it will take uncountable days to sync to 100%. So I tried to start with a snapshot.

This is how I start it:

#!/bin/bash

DOCKER_BUILDKIT=1 \
COMPOSE_DOCKER_CLI_BUILD=1 \
CHAIN_FOLLOWER_START_SLOT=23068800 \
CHAIN_FOLLOWER_START_ID=a650a3f398ba4a9427ec8c293e9f7156d81fd2f7ca849014d8d2c1156c359b3a \
RESTORE_SNAPSHOT=https://update-cardano-mainnet.iohk.io/cardano-db-sync/13.4/db-sync-snapshot-schema-13.4-block-10717727-x86_64.tgz \
docker compose up -d --build

docker-compose.yml (just cardano-db-syn part)

  cardano-db-sync:
    platform: linux/x86_64
    image: ghcr.io/intersectmbo/cardano-db-sync:${CARDANO_DB_SYNC_VERSION:-13.3.0.0}
    command: [
      "--config", "/config/cardano-db-sync/config.json",
      "--socket-path", "/node-ipc/node.socket"
    ]
    environment:
      - POSTGRES_HOST=postgres
      - POSTGRES_PORT=${POSTGRES_PORT:-5432}
      - RESTORE_SNAPSHOT=${RESTORE_SNAPSHOT:-}
      - RESTORE_RECREATE_DB=Y
    depends_on:
      - cardano-node
      - postgres
    secrets:
      - postgres_password
      - postgres_user
      - postgres_db
    volumes:
      - ./config/network/${NETWORK:-mainnet}:/config
      - /workspace/db-sync-sync/:/var/lib/cexplorer
      - /workspace/cardano-node-sync/socket:/node-ipc
    restart: on-failure
    stop_signal: SIGINT
    logging:
      driver: "json-file"
      options:
        max-size: "200k"
        max-file: "10"

Error from cardano-db-sync container:

Downloading snapshot https://update-cardano-mainnet.iohk.io/cardano-db-sync/13.4/db-sync-snapshot-schema-13.4-block-10717727-x86_64.tgz
db-sync-snapshot-schema-13.4-block-10717727-x86_64.tgz: OK
ERROR:  cannot drop objects owned by role graphql because they are required by the database system

What do I have to do to ensure that the sync runs smoothly and I don’t get any more errors?