Thin Volumes and Snapshot Retention Pressure on a Full Pool
A PowerStore 1200T has one pool with 80 TB usable capacity. It hosts 20 thin volumes of 5 TB each for a mix of SQL and file workloads; hosts have written 60% of the logical capacity. A protection policy runs daily snapshots on all 20 volumes and retains 14 copies. The pool is at 78% used, and the daily change rate is about 5% of written data. A new project requests a 10 TB volume with 30-day snapshot retention and asks whether it can be created immediately. The storage team also wants to know whether the existing snapshot policy is safe. There is no replication configured, and the array has one expansion shelf available but not installed. Peak write latency is normal.
1. Can the 10 TB volume be created? Explain thin provisioning and the capacity risk.
Read the worked answer
Yes, the 10 TB thin volume can be created because thin provisioning does not reserve 10 TB of physical capacity. That does not mean it is safe. The pool has about 17.6 TB free (22% of 80 TB), and the new volume will consume space as data is written. If the host writes the full 10 TB, free space drops to roughly 7.6 TB before any snapshots. The request should be approved only with a capacity plan: confirm expected written capacity, set a snapshot policy that fits the free space, and monitor pool utilization. Creation is a provisioning action; the judgement is whether the capacity and protection policy are sustainable.
2. Estimate the snapshot space pressure and recommend a protection policy change.
Read the worked answer
Existing written data is about 60 TB. At a 5% daily change rate, that is about 3 TB/day of changed blocks. Fourteen daily snapshots could consume up to roughly 42 TB if every change were unique; dedupe and compression and snapshot sharing reduce that, but the pool is already at 78%. A 30-day daily retention on the new volume adds more pressure. Replace the blanket 14-daily policy with a tiered rule, for example 7 daily plus 4 weekly, and apply it selectively. Keep longer retention only where required. Install or expand capacity before adding 30-day retention, and monitor snapshot space separately from volume space.
3. What operational guardrails would you set before installing the expansion shelf?
Read the worked answer
Before installing the shelf, confirm rack space, power, and back-end cabling, and verify that the array supports the additional drive count. More importantly, set guardrails: capacity alerts at 70%, 80%, and 90%; a snapshot space alert; and a review of protection policies. Do not use the shelf as an excuse to keep 30-day daily snapshots on every volume. After expansion, recheck pool balance and performance; adding capacity does not add node CPU, so if latency is already rising, validate performance before assuming the shelf solves it.