Update docs/running-at-massive-scale.md

Co-authored-by: Yuan Tang <terrytangyuan@gmail.com>

docs/running-at-massive-scale.md

@@ -17,7 +17,7 @@ too small. We recommend you test your cluster to make sure it can run the number
 installing Argo. Create pods at the rate you expect that it'll be created in production. Make sure Kubernetes can keep
 up with requests to delete pods at the same rate.
 
-You'll need to to GC data quickly. The less data that Kubernetes and Argo deal with, the less work they need to do. Use
+You'll need to GC data quickly. The less data that Kubernetes and Argo deal with, the less work they need to do. Use
 pod GC and workflow GC to achieve this.
 
 ## Overwhelmed Kubernetes API
@@ -36,4 +36,3 @@ for 5m. You can reduce the number of records create by setting `DEFAULT_REQUEUE_
 but will suit workflows where nodes run for over 1m.
 
 
-