doc: Add newlines after headers.

Some Markdown parsers require this newline. Our remaining files always
have this newline. Let's use it here as well to be consistent.

doc/HorizontalReshardingWorkflowGuide.md

@@ -1,10 +1,10 @@
-
 This guide shows you an example about how to apply range-based sharding
 process in an existing unsharded Vitess [keyspace](http://vitess.io/overview/concepts.html#keyspace)
 using the horizontal resharding workflow. In this example, we will reshard
 from 1 shard "0" into 2 shards "-80" and "80-".
 
 ## Overview
+
 The horizontal resharding process mainly contains the following steps
 (each step is a phase in the workflow):
 
@@ -73,7 +73,9 @@ the resharding process:
     "Reset Job". Otherwise, the vtworker is not ready for executing other tasks.
 
 ## Horizontal resharding workflow
+
 ### Create the workflow
+
 1.  Open the *Workflows* section on the left menu of vtctld UI (http://localhost:15000).
     Click the "+" button in the top right corner to open the "Create
     a new Workflow" dialog.
@@ -113,6 +115,7 @@ to 2 destination shards, therefore you can see task node "Shard -80" and
 [this](https://cloud.githubusercontent.com/assets/23492389/24313539/71c9c8ae-109a-11e7-9e4a-0c3e8ee8ba85.png). 
 
 ### Approvals of Tasks Execution (Canary feature)
+
 Once the workflow start to run (click the "Start" button if you selected
 "Skip Start" and the workflow hasn't started yet.), you need to approve the
 task execution for each phase if "enable_approvals" is selected. The approvals
@@ -135,6 +138,7 @@ approval. But you don't need to approve the same tasks again for a restarted
 workflow.
 
 ### Retry
+
 A "Retry" button will be enabled under the task node if the task failed (click
 the task node if your job get stuck but don't see the Retry button). Click this
 button if you have fixed the bugs and want to retry the failed task. You can
@@ -153,6 +157,7 @@ that runs tasks sequentially, remaining unstarted tasks under this phase will
 no long be executed. The phases afterwards will no longer be executed.
 
 ### Checkpoint and Recovery
+
 The resharding workflow tracks the status for every task and checkpoint these
 status into topology server whenever there is a status update. When a workflow
 is stopped and restarted by loading the checkpoint in the topology, it can
@@ -160,6 +165,7 @@ continue to run all the unfinished tasks.
 
 
 ## Verify Results and Clean up
+
 After the resharding process, data in the original shard is identically copied
 to new shards. Data updates will be visible on the new shards, but not on the
 original shard. You should then see in the vtctld UI *Dashboard* page that shard
@@ -219,10 +225,12 @@ vitess/examples/local$ ./zk-down.sh
 ```
 
 ## Reference
+
 You can checkout the old version tutorial [here](http://vitess.io/user-guide/horizontal-sharding.html).
 It walks you through the resharding process by manually executing commands.
 
 ### Details in SplitClone phase
+
 *vtworker* copies data from a paused snapshot. It will pause replication on
 one rdonly (offline processing) tablet to serve as a consistent snapshot of
 the data. The app can continue without downtime, since live traffic is served
@@ -238,6 +246,7 @@ range, so it identifies -80 and 80- as the destination shards, since they
 combine to cover the same range.
 
 ### Details in WaitForFilteredReplication phase
+
 Once the copying from a paused snapshot (phase SplitClone) has finished,
 *vtworker* turns on [filtered replication](http://vitess.io/user-guide/sharding.html#filtered-replication),
 which allows the destination shards to catch up on updates that have continued
@@ -245,6 +254,7 @@ to flow in from the app since the time of the snapshot. After the destination
 shards are caught up, they will continue to replicate new updates. 
 
 ### Details in MigrateServedTypeMaster phase
+
 During the *master* migration, the original shard masters will first stop
 accepting updates. Then the process will wait for the new shard masters to
 fully catch up on filtered replication before allowing them to begin serving.

doc/ShardingKubernetesWorkflow.md

@@ -8,6 +8,7 @@ except that here we'll give you the commands you'll need in the kubernetes
 environment.
 
 ## Overview
+
 The horizontal resharding process overview can be found
 [here](http://vitess.io/user-guide/horizontal-sharding-workflow.html#overview) 
 
@@ -68,7 +69,9 @@ resharding process:
     "Reset Job". Otherwise, the vtworker is not ready for executing other tasks.
 
 ## Horizontal Resharding Workflow
+
 ### Create the Workflow
+
 Using the web vtctld UI to create the workflow is the same with [steps in local
 environment](http://vitess.io/user-guide/horizontal-sharding-workflow.html#create-the-workflow)
 except for filling the "vtworker Addresses" slot. You need to get the external
@@ -83,18 +86,22 @@ vitess/examples/kubernetes$ ./kvtctl.sh WorkflowCreate -skip_start=false horizon
 ```
 
 ### Approvals of Tasks Execution (Canary feature)
+
 Please check the content in general 
 [Horizontal Sharding guide](http://vitess.io/user-guide/horizontal-sharding-workflow.html#approvals-of-tasks-execution-canary-feature)
 
 ### Retry
+
 Please check the content in general 
 [Horizontal Sharding guide](http://vitess.io/user-guide/horizontal-sharding-workflow.html#retry)
 
 ### Checkpoint and Recovery
+
 Please check the content in general 
 [Horizontal Sharding guide](http://vitess.io/user-guide/horizontal-sharding-workflow.html#checkpoint-and-recovery)
 
 ## Verify Results and Clean up
+
 After the resharding process, data in the original shard is identically copied
 to new shards. Data updates will be visible on the new shards, but not on the
 original shard. You should then see in the vtctld UI *Dashboard* page that shard
@@ -170,6 +177,7 @@ $ gcloud compute firewall-rules delete vtctld guestbook
 ```
 
 ## Reference
+
 You can checkout the old version tutorial [here](http://vitess.io/user-guide/sharding-kubernetes.html).
 It walks you through the resharding process by manually executing commands.
 
@@ -178,7 +186,9 @@ kubernetes cluster, you can check out more information
 [here](https://kubernetes.io/docs/user-guide/kubectl-overview).
 
 ## Troubleshooting
-### Checking status of your setup. 
+
+### Checking status of your setup.
+
 To get status of pods and services you've setup, you can use the commands
 (all pods should be in Running status, guestbook and vtworker services
 should have assign external IP):
@@ -189,6 +199,7 @@ $ kubectl get services
 ```
 
 ### Debugging pods.
+
 If you find out a component (e.g. vttablet, vtgate) doesn't respond as
 expected, you can surface the log using this command (the pod name can be
 found out using the command mentioned above):
@@ -201,6 +212,7 @@ $ kubectl logs <pod name> [-c <container>]
 ```
 
 ### Debugging pending external IP issue.
+
 If you found that your service has a pending external IP for long time, it
 maybe because you've reached the limitation of networking resource. Please
 go to your project console on gcloud (cloud.google.com), then go to *Load