Bucket health mode should be calculated according to mirror storage

[dannyzaken]

Environment info

• Version: 2.1
• Deployment: AZURE | AWS | GCLOUD | ESX | VBOX | DEV
• Customer: NAME

Actual behavior

1. currently the bucket health is calculated by checking all pools in all tiers and checking if pools are valid in the tiering_policy_status

Expected behavior

1. a better and more consistent way to calculate the validity is by looking at the available storage in the mirror set, which is also returned in the tiering_policy_status.
2. the available storage in the mirror set already takes into account the number of blocks required for chunk vs the amount of storage in each node in the pools set.
3. when the available storage is greater than 0 than the mirror has enough healthy nodes