link fixing

Author: tomvcassidy

articles/service-fabric/service-fabric-cluster-resource-manager-balancing.md

@@ -78,7 +78,7 @@ The Cluster Resource Manager also needs some additional information to determine
 ## Balancing thresholds
 A Balancing Threshold is the main control for triggering rebalancing. The Balancing Threshold for a metric is a _ratio_. If the load for a metric on the most loaded node divided by the amount of load on the least loaded node exceeds that metric's *BalancingThreshold*, then the cluster is imbalanced. As a result balancing is triggered the next time the Cluster Resource Manager checks. The *MinLoadBalancingInterval* timer defines how often the Cluster Resource Manager should check if rebalancing is necessary. Checking doesn't mean that anything happens. 
 
-Balancing Thresholds are defined on a per-metric basis as a part of the cluster definition. For more information on metrics, check out [this article](service-fabric-cluster-resource-manager-metrics.md).
+Balancing Thresholds are defined on a per-metric basis as a part of the cluster definition. For more information on metrics, check out the [metrics article](service-fabric-cluster-resource-manager-metrics.md).
 
 ClusterManifest.xml
 
@@ -193,8 +193,8 @@ As we described in the earlier sections, the main controls of triggering rebalan
 ### How balancing per node type affects a cluster
 
 During balancing of a cluster per node type, the Service Fabric Cluster Resource Manager calculates the imbalance state for each node type. If at least one node type is imbalanced, the balancing phase will be triggered. Balancing phase won't move replicas on node types that are imbalanced, when balancing is temporarily paused on these node types (e.g. minimal balancing interval hasn't passed since a previous balancing phase). The detection of an imbalanced state uses common mechanisms already available for classical cluster balancing, but improves configuration granularity and flexibility. The mechanisms used for balancing per node type to detect imbalance are provided in the list below:
-- **Metric balancing thresholds** per node type are values that have a similar role as the globally defined balancing threshold used in classical balancing. The ratio of minimum and maximum metric load is calculated for each node type. If that ratio of a node type is higher than the defined balancing threshold of the node type, the node type is marked as imbalanced. For more details regarding configuration of metric activity thresholds per node type, please check [here](service-fabric-cluster-resource-manager-balancing#balancing-thresholds-per-node-type).
-- **Metric activity thresholds** per node type are values that have a similar role to the globally defined activity threshold used in classical balancing. The maximum metric load is calculated for each node type. If the maximum load of a node type is higher than the defined activity threshold for that node type, the node type is marked as imbalanced. For more details regarding configuration of metric activity thresholds per node type, please check [here](service-fabric-cluster-resource-manager-balancing#activity-thresholds-per-node-type).
+- **Metric balancing thresholds** per node type are values that have a similar role as the globally defined balancing threshold used in classical balancing. The ratio of minimum and maximum metric load is calculated for each node type. If that ratio of a node type is higher than the defined balancing threshold of the node type, the node type is marked as imbalanced. For more details regarding configuration of metric activity thresholds per node type, please check the [balancing thresholds per node type section](#balancing-thresholds-per-node-type).
+- **Metric activity thresholds** per node type are values that have a similar role to the globally defined activity threshold used in classical balancing. The maximum metric load is calculated for each node type. If the maximum load of a node type is higher than the defined activity threshold for that node type, the node type is marked as imbalanced. For more details regarding configuration of metric activity thresholds per node type, please check the [activity-thresholds-per-node-type section](#activity-thresholds-per-node-type).
 - **Minimum balancing interval** per node type has a role similar to the globally defined minimum balancing interval. For each node type, the Cluster Resource Manager preserves the timestamp of the last balancing. Two consecutive balancing phases couldn't be executed on a node type within the defined minimum balancing interval. For more details regarding configuration of minimum balancing interval per node type, please check the [minimum balancing interval per node type section](#minimum-balancing-interval-per-node-type).
 
 ### Describing balancing per node type
@@ -317,7 +317,7 @@ Customer detected that workloads of these node types have different balancing ne
 Maximum load of node type **A** of *600* is lower than defined activity threshold of *700*, thus node type **A** won't be balanced. Maximum load of node type **B** of *900* is higher than defined activity threshold of *200*. The node type **B** violates activity threshold criteria. Maximum load of node type **C** of *600* is higher than defined activity threshold of *300*. The node type **C** violates activity threshold criteria. The node type **B** doesn't violate balancing threshold criteria, since current ratio of maximum and minimum load for this node type is *9*, but balancing threshold is *10*. Node type **C** violates balancing threshold criteria, since current ratio of maximum and minimum load is *2*, and balancing threshold is *2*.  Balancing is required only for replicas in the node type **C**, and the only set of replicas that will be eligible for movements during balancing phase are replicas placed in the node type **C**. 
 
 ## Next steps
-* Metrics are how the Service Fabric Cluster Resource Manger manages consumption and capacity in the cluster. To learn more about metrics and how to configure them, check out [this article](service-fabric-cluster-resource-manager-metrics.md)
-* Movement Cost is one way of signaling to the Cluster Resource Manager that certain services are more expensive to move than others. For more about movement cost, refer to [this article](service-fabric-cluster-resource-manager-movement-cost.md)
-* The Cluster Resource Manager has several throttles that you can configure to slow down churn in the cluster. They're not normally necessary, but if you need them you can learn about them [here](service-fabric-cluster-resource-manager-advanced-throttling.md)
-* The Cluster Resource Manager can recognize and handle subclustering (a situation that sometimes arises when you use placement constraints and balancing). To learn how subclustering can affect balancing and how you can handle it, see [here](cluster-resource-manager-subclustering.md)
+* Metrics are how the Service Fabric Cluster Resource Manger manages consumption and capacity in the cluster. To learn more about metrics and how to configure them, check out the [metrics article](service-fabric-cluster-resource-manager-metrics.md)
+* Movement Cost is one way of signaling to the Cluster Resource Manager that certain services are more expensive to move than others. For more about movement cost, refer to the [movement cost article](service-fabric-cluster-resource-manager-movement-cost.md)
+* The Cluster Resource Manager has several throttles that you can configure to slow down churn in the cluster. They're not normally necessary, but if you need them you can learn about them the [advanced throttling article](service-fabric-cluster-resource-manager-advanced-throttling.md)
+* The Cluster Resource Manager can recognize and handle subclustering. Subclustering can arise when you use placement constraints and balancing. To learn how subclustering can affect balancing and how you can handle it, see the [subclustering article](cluster-resource-manager-subclustering.md)