Plan UniFi Protect storage and retention

Goal and requirements

Estimate storage before buying drives, then verify retention using the installed system. You need a camera inventory, intended recording mode, retention target and the supported storage configuration for your recorder. Use an average recording bitrate for each camera. Resolution alone does not determine the amount of data written.

Write the quality requirement alongside the duration, for example, 30 days at the selected recording quality. A target expressed only as days can hide a change to lower-quality older footage. Protect's Enhanced Retention can reduce the quality of older recordings, and that reduction cannot subsequently be reversed. [1]

Calculate the starting estimate

For continuous recording, using decimal units:

Required GB = total average bitrate in Mb/s × 10.8 × days

Required TB = required GB ÷ 1,000

The factor 10.8 follows from 86,400 seconds per day, eight bits per byte and decimal unit conversions. This calculation estimates video payload. It does not promise actual retention or include every filesystem, database, audio, container and operational overhead. Variable scene activity also changes average bitrate.

Consider an explicitly assumed example: four cameras average 6 Mb/s each and record continuously for 30 days. The total is 24 Mb/s. The estimate is 24 × 10.8 × 30 = 7,776 GB, or 7.776 TB. Adding an illustrative planning allowance of 20% gives 9.331 TB. That allowance is a budgeting choice, not a Ubiquiti retention guarantee.

Compare raw and usable capacity

Four 4 TB drives provide 16 TB of raw labelled capacity. With a conventional four-drive RAID 5 arrangement, the simple nominal usable estimate is approximately 12 TB before formatting and system overhead. RAID 10 with four equal drives provides approximately 8 TB before overhead. Use the capacity and protection mode actually reported by the supported recorder configuration. Ubiquiti offers different protection options depending on recorder and drive count. [2]

In the assumed example, the nominal 12 TB option exceeds the 9.331 TB planning figure, while the nominal 8 TB option does not. This is an arithmetic comparison, not a recommendation to sacrifice redundancy. Select a recorder and supported disks that satisfy both your usable-capacity requirement and your drive-failure protection requirement.

RAID helps maintain data availability through specified drive failures. An independent copy addresses a different problem, such as loss of the whole recorder. Keep footage that must survive recorder loss in a separate export or archive.

Configure and measure

  1. Inventory every stream. Create a table with camera name, recording mode, selected quality, estimated average Mb/s and target days. Include planned additional cameras. Where bitrate information is unavailable, label the value assumed and plan to replace it with a measurement.
  2. Calculate each scenario. Sum average bitrates and apply the formula. Repeat for the expected future camera count. Do not assume Detection Only consumes a fixed percentage of continuous recording. Traffic, weather and detector settings determine how much footage is recorded, so measure that configuration separately.
  3. Confirm supported capacity. Check the exact recorder's supported drive types, bay count and storage mode. Open its storage page and record actual usable capacity, health and any rebuild status. A number printed on a disk box is not the recording capacity of the assembled array. [2]
  4. Set a clear recording policy. Use Recording Manager in Protect 7.2 and later; older versions also expose settings in individual camera panels. Select the quality and recording mode used in your calculation. In Storage Manager, leave Enhanced Retention off during a full-quality baseline unless reduced-quality older footage is intended. Record the encoding setting too: Protect 7.3's encoding and retention options can change storage use. [1] [3] [4]
  5. Record a representative period. Allow normal daytime and night-time use. Include busy periods relevant to the location. Record storage consumption and the age of the oldest playable footage each day. For a newly initialised system, several days of data can reveal a poor estimate, but reaching a 30-day retention target requires observing an adequate interval.
  6. Reconcile the estimate. If a representative day writes 300 GB, a 30-day payload estimate is 9 TB before additional allowances. Compare that with available recording capacity. Once overwriting begins, verify the oldest accessible footage rather than dividing total disk size by a brief quiet-period bitrate.
  7. Adjust one factor. Options include more supported usable storage, fewer retained days, a different recording mode or an accepted quality change. Recalculate after the change. If enabling Enhanced Retention, document both the high-quality period and the older low-quality period separately. [1]

Verify and recover

Play a clip near the oldest expected date from every important camera. Check that it has the promised quality and that there are no unexplained recording gaps. Export important footage before storage changes.

If the array is degraded, investigate the identified drive and follow the recorder's replacement procedure. Avoid experimenting with storage protection modes on the only copy of required recordings. Record the last known working configuration and obtain independent exports before any operation that warns of data deletion.

Sources

Back to the series

Leave a comment

Please note, comments need to be approved before they are published.