RockteMQ-AI opened a new pull request, #10516:
URL: https://github.com/apache/rocketmq/pull/10516
## Motivation
When a TCP connection is broken without a RST, the gRPC client needs to rely
on HTTP/2 PING (keepalive) to detect the dead connection. Currently, the gRPC
server in the Proxy does **not** configure `permitKeepAliveTime` or
`permitKeepAliveWithoutCalls`, which means gRPC Netty server defaults apply:
- `permitKeepAliveTime` = **5 minutes** — clients cannot send keepalive
pings more frequently than every 5 minutes
- `permitKeepAliveWithoutCalls` = **false** — keepalive pings on idle
connections are rejected
This causes slow dead-connection detection (up to 5.5 minutes) and
ineffective idle connection health checks.
Fixes #10510
## Changes
Add two configurable parameters to `ProxyConfig`:
| Parameter | Default | Description |
|-----------|---------|-------------|
| `grpcServerPermitKeepAliveTimeMillis` | 10000 (10s) | Minimum time between
client keepalive pings |
| `grpcServerPermitKeepAliveWithoutCalls` | true | Allow keepalive pings
when no active RPCs |
Apply these in `GrpcServerBuilder`:
```java
serverBuilder
.permitKeepAliveTime(config.getGrpcServerPermitKeepAliveTimeMillis(),
TimeUnit.MILLISECONDS)
.permitKeepAliveWithoutCalls(config.isGrpcServerPermitKeepAliveWithoutCalls());
```
## Impact
- **Backward compatible**: Default values are more permissive than current
implicit defaults
- **Minimal overhead**: ~8KB/hour per connection at 30s interval
- **Enables faster detection**: Once clients reduce keepAliveTime,
dead-connection detection can drop from 5.5 minutes to ~40 seconds
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
