Scaling GraphQL subscritpion to 1 Million subscrptions using Hasura GraphQL Engine and YugabyteDB deployed on Kubernetes
Results obtained by running the benchmarks on a kubernetes cluster where GraphQL subscriptions are scaled from 25K subscriptions to 1 Million subscriptions linearly without any downtime.
| No. of Subscriptions | YugabyteDB: Number of vCPUs | Hasura: Number of vCPUs |
|---|---|---|
| 25K | 8 | 4 |
| 50K | 16 | 12 |
| 100k | 32 | 20 |
| 1M | 320 | 200 |
| No. of Subscriptions | YugabyteDB Cluster | Hasura Pods | GraphQL Subscription Tool |
|---|---|---|---|
| 25K | 4 pods, 2 vcpus (4GB RAM, 2x50GB SSD) | 1 Pod * 4 vCPUs, 8GB RAM | 25k subscribers, 4vcpu, 8GB RAM |
| 50K | 4 Tservers - 4 vCPUs/8 GB, 2 * 50 GB SSD | 3*pod * 4xCPU, 8GB RAM | 50k subscribers, 4vcpu, 8GB RAM |
| 100k | 4 Tservers - 8 vCPU, 16 GB, 2 * 50 GB SSD | 5 pod * 4vcpu/ 8GB RAM | 100K subscribers 4vcpu, 8GB RAM |
| 1M | 20 pods of 16 vCPUs each | 50 Pod * 4 Vcpu, 8 GB RAM | 1M subscribers 8vcpu, 16GB RAM |
- Live queries (subscriptions) latency.
- YugabyteDB CPU utilization during GraphQL client subscription initilization.
- Hasura CPU utilization.
