mirror of
https://github.com/QuilibriumNetwork/ceremonyclient.git
synced 2026-02-23 11:27:25 +08:00
* v2.1.0 [omit consensus and adjacent] - this commit will be amended with the full release after the file copy is complete * 2.1.0 main node rollup
82 lines
2.0 KiB
Go
82 lines
2.0 KiB
Go
package internal
|
|
|
|
import (
|
|
"context"
|
|
"sync"
|
|
"time"
|
|
|
|
"github.com/libp2p/go-libp2p/core/host"
|
|
"github.com/libp2p/go-libp2p/core/peer"
|
|
"github.com/libp2p/go-libp2p/p2p/protocol/ping"
|
|
"go.uber.org/zap"
|
|
)
|
|
|
|
type peerMonitor struct {
|
|
ps *ping.PingService
|
|
timeout time.Duration
|
|
period time.Duration
|
|
attempts int
|
|
}
|
|
|
|
func (pm *peerMonitor) pingOnce(ctx context.Context, logger *zap.Logger, peer peer.ID) bool {
|
|
pingCtx, cancel := context.WithTimeout(ctx, pm.timeout)
|
|
defer cancel()
|
|
select {
|
|
case <-ctx.Done():
|
|
case <-pingCtx.Done():
|
|
logger.Debug("ping timeout")
|
|
return false
|
|
case res := <-pm.ps.Ping(pingCtx, peer):
|
|
if res.Error != nil {
|
|
logger.Debug("ping error", zap.Error(res.Error))
|
|
return false
|
|
}
|
|
logger.Debug("ping success", zap.Duration("rtt", res.RTT))
|
|
}
|
|
return true
|
|
}
|
|
|
|
func (pm *peerMonitor) ping(ctx context.Context, logger *zap.Logger, wg *sync.WaitGroup, peer peer.ID) {
|
|
defer wg.Done()
|
|
for i := 0; i < pm.attempts; i++ {
|
|
pm.pingOnce(ctx, logger, peer)
|
|
}
|
|
}
|
|
|
|
func (pm *peerMonitor) run(ctx context.Context, logger *zap.Logger) {
|
|
for {
|
|
select {
|
|
case <-ctx.Done():
|
|
return
|
|
case <-time.After(pm.period):
|
|
peers := pm.ps.Host.Network().Peers()
|
|
logger.Debug("pinging connected peers", zap.Int("peer_count", len(peers)))
|
|
wg := &sync.WaitGroup{}
|
|
for _, id := range peers {
|
|
slogger := logger.With(zap.String("peer_id", id.String()))
|
|
wg.Add(1)
|
|
go pm.ping(ctx, slogger, wg, id)
|
|
}
|
|
wg.Wait()
|
|
logger.Debug("pinged connected peers")
|
|
}
|
|
}
|
|
}
|
|
|
|
// MonitorPeers periodically looks up the peers connected to the host and pings them
|
|
// repeatedly to ensure they are still reachable. If the peer is not reachable after
|
|
// the attempts, the connections to the peer are closed.
|
|
func MonitorPeers(
|
|
ctx context.Context, logger *zap.Logger, h host.Host, timeout, period time.Duration, attempts int,
|
|
) {
|
|
ps := ping.NewPingService(h)
|
|
pm := &peerMonitor{
|
|
ps: ps,
|
|
timeout: timeout,
|
|
period: period,
|
|
attempts: attempts,
|
|
}
|
|
|
|
go pm.run(ctx, logger)
|
|
}
|