roachtest: set max lifetime for connections in local runs

This change fixes an issue observed when running roachtests that
restart nodes locally. This was primarily observed on M1s, but we
apply the same change regardless of platform for consistency.

Epic: none

Release note: None

Author: Renato Costa

pkg/cmd/roachtest/cluster.go

@@ -2677,6 +2677,22 @@ func (c *clusterImpl) ConnE(
 	if err != nil {
 		return nil, err
 	}
+
+	// When running roachtest locally, we set a max connection lifetime
+	// to avoid errors like the following:
+	//
+	// `read tcp 127.0.0.1:63742 -> 127.0.0.1:26257: read: connection reset by peer`
+	//
+	// The pq issue below seems related. This was only observed in local
+	// runs so the lifetime is only applied in that context intentionally;
+	// for cloud runs, we use the connection pool's default behaviour.
+	//
+	// https://github.com/lib/pq/issues/835
+	if c.spec.Cloud == spec.Local {
+		localConnLifetime := 10 * time.Second
+		db.SetConnMaxLifetime(localConnLifetime)
+	}
+
 	return db, nil
 }