Skip to content

DELETE_AND_REDEPLOY_SUBGRAPH.md

Latest commit

 

History

History
181 lines (129 loc) · 5.06 KB

DELETE_AND_REDEPLOY_SUBGRAPH.md

File metadata and controls

181 lines (129 loc) · 5.06 KB

Delete and Redeploy Subgraph - Dashboard Method

Overview

Since the IPFS hash isn't accessible via CLI and the subgraph has errors, the easiest approach is to delete the broken subgraph from the Goldsky dashboard and deploy a fresh one.

Step 1: Delete Broken Subgraph

  1. Go to Goldsky Dashboard:

  2. Find the Subgraph:

    • Go to the "Subgraphs" section
    • Find metokens/v0.0.1 (the one with the block hash mismatch error)

  3. Delete the Subgraph:

    • Click on the subgraph to open its details
    • Look for a "Delete" or "Remove" button (usually in settings or actions menu)
    • Confirm the deletion

Note: This will remove the broken subgraph. Historical data will be lost, but since it's broken anyway, this is fine.

Step 2: Deploy New Subgraph via Dashboard

Option A: Deploy as v0.0.1 (Same Version)

If you want to keep the same version name:

  1. In Goldsky Dashboard:

    • Go to "Subgraphs" section
    • Click "New Subgraph" or "Deploy Subgraph"

  2. Configure Deployment:

    • Name: metokens
    • Version: v0.0.1
    • Network: base (Base Mainnet)
    • Source: You'll need to provide:
      • IPFS hash: QmVaWYhk4HKhk9rNQi11RKujTVS4KHF1uHGNVUF4f7xJ53
      • OR upload subgraph files if you have them locally
      • OR use the deployment wizard if available

  3. Deploy:

    • Click "Deploy" or "Create"
    • Wait for deployment to start

Option B: Deploy as v0.0.2 (New Version) - Recommended

Since the pipeline is already configured for v0.0.2, this is the recommended approach:

  1. In Goldsky Dashboard:

    • Go to "Subgraphs" section
    • Click "New Subgraph" or "Deploy Subgraph"

  2. Configure Deployment:

    • Name: metokens
    • Version: v0.0.2
    • Network: base (Base Mainnet)
    • Source: Provide IPFS hash or upload files

  3. Deploy:

    • Click "Deploy" or "Create"
    • Wait for deployment to start

Step 3: Verify New Subgraph

After deployment, verify the subgraph is working:

# Check subgraph status
goldsky subgraph list metokens/v0.0.2

# Test the endpoint
curl -X POST https://api.goldsky.com/api/public/project_cmh0iv6s500dbw2p22vsxcfo6/subgraphs/metokens/v0.0.2/gn \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ subscribes(first: 1) { id meToken hubId blockTimestamp } }"
  }'

Expected Response:

{
  "data": {
    "subscribes": [...]
  }
}

If you get {"errors":[{"message":"indexing_error"}]}, wait a bit for indexing to start.

Step 4: Deploy Mirror Pipeline

Once the subgraph is healthy and indexing, deploy the Mirror pipeline:

# Validate pipeline (already configured for v0.0.2)
goldsky pipeline validate pipeline-metokens-all.yaml

# Deploy the pipeline
goldsky pipeline apply pipeline-metokens-all.yaml --status ACTIVE

Alternative: If Dashboard Doesn't Have Deploy Option

If the dashboard doesn't allow deploying from IPFS hash, you have two options:

Option 1: Contact Goldsky Support

Ask them to:

  1. Delete the broken metokens/v0.0.1 subgraph
  2. Deploy a fresh metokens/v0.0.2 from IPFS hash QmVaWYhk4HKhk9rNQi11RKujTVS4KHF1uHGNVUF4f7xJ53

Email Template:

Subject: Request to Delete and Redeploy MeTokens Subgraph

Hello Goldsky Support,

I need assistance with my MeTokens subgraph:

1. Delete the broken subgraph: metokens/v0.0.1
   - Deployment ID: QmVaWYhk4HKhk9rNQi11RKujTVS4KHF1uHGNVUF4f7xJ53
   - It has a block hash mismatch error and is not functioning

2. Deploy a fresh subgraph: metokens/v0.0.2
   - From the same IPFS hash: QmVaWYhk4HKhk9rNQi11RKujTVS4KHF1uHGNVUF4f7xJ53
   - Network: Base Mainnet
   - Project: project_cmh0iv6s500dbw2p22vsxcfo6

This is needed to fix the indexing errors and enable our Mirror pipeline.

Thank you!

Option 2: Use CLI After Dashboard Deletion

  1. Delete via dashboard (as described in Step 1)
  2. Try CLI deployment again (might work after deletion): 
    goldsky subgraph deploy metokens/v0.0.2 \
  --from-ipfs-hash QmVaWYhk4HKhk9rNQi11RKujTVS4KHF1uHGNVUF4f7xJ53 \
  --description "Fresh deployment for Mirror pipeline"

Current Pipeline Configuration

The pipeline-metokens-all.yaml is already configured for v0.0.2:

sources:
  metoken_subscribes_source:
    name: subscribe
    subgraphs:
      - name: metokens
        version: v0.0.2  # ← Ready for new deployment

If you deploy as v0.0.1 instead, you'll need to update the pipeline back to v0.0.1.

Summary

  1. ✅ Delete broken metokens/v0.0.1 from dashboard
  2. ✅ Deploy fresh metokens/v0.0.2 from dashboard (or contact support)
  3. ✅ Wait for subgraph to start indexing
  4. ✅ Verify subgraph is working
  5. ✅ Deploy Mirror pipeline (already configured for v0.0.2)

Benefits

  • ✅ Clean slate - no corrupted state
  • ✅ Fresh indexing from start
  • ✅ No block hash mismatch errors
  • ✅ Pipeline already configured and ready

Last Updated: 2025-01-20
Method: Dashboard deletion + redeployment
New Version: v0.0.2 (recommended) or v0.0.1 (if preferred)