Subgraph Rollout Process

Checklist

• Create a coordination slack thread for the development team.
  • Including a complete matrix of what vendor, network combination to track.
• Create three workflow runs:
  • with parameters

    from: tags/subgraph release tag
    vendor: superfluid|graph|satsuma
    deployment_env: v1
    network: `all` | "hand picked network"
    

  • Send a slack message.
• QA to validate from console that all deployment has matching sfmeta.packageVersion & sfmeta.id (github commit hash)
• Mark rollout complete.
• Do a retro of the rollout, and see what could be improved next time around.

Known Issues

1. Problematic endpoints:

• Superfluid-hosted polygon (using alchemy archive node, expensive)
• Some graph-hosted endpoints may take long: CELO

2. Primary and fallback may be out of sync for a while.