fix: don't immediately schedule monitoring after streaming failure

When a failure occurs while streaming topology updates, we need to
wait some time before restarting the monitoring protocol. It's
likely that if some error closed the connection, an immediate
attempt to reconnect will also fail, leading to tight failure
loops.

NODE-2711

Author: mbroadst

src/sdam/monitor.ts

@@ -137,8 +137,7 @@ class Monitor extends EventEmitter {
     const minHeartbeatFrequencyMS = this.options.minHeartbeatFrequencyMS;
     this[kMonitorId] = makeInterruptableAsyncInterval(monitorServer(this), {
       interval: heartbeatFrequencyMS,
-      minInterval: minHeartbeatFrequencyMS,
-      immediate: true
+      minInterval: minHeartbeatFrequencyMS
     });
   }