JuniorStorage

How does object storage achieve such high durability?

What they are really testing: Whether you know it stores redundant copies (or erasure-coded shards) across failure domains and uses checksums to detect and repair corruption, rather than a single copy on one disk.

A real interview question

How does object storage achieve such high durability?

What most people say

drag me

It keeps the data on reliable disks in the data center.

Reliable disks alone do not give eleven nines, disks fail constantly at scale. Durability comes from redundancy across failure domains (copies or erasure coding), checksums for corruption, and automatic self-healing. The mechanism is the answer.

The follow-ups they ask next

  • What is erasure coding and why use it over full replication?

    The object is split into data + parity shards; any sufficient subset reconstructs the original. It gives high durability with much less storage overhead than keeping several full copies, the trade-off being reconstruction compute on read/repair.

  • How does the system handle silent corruption (bit rot)?

    Checksums on stored data plus continuous scrubbing detect corruption, and the bad copy/shard is repaired from healthy redundant copies. Without integrity checks, corruption would silently spread.

What the interviewer is listening for

  • Redundancy across failure domains
  • Knows erasure coding
  • Checksums + self-healing repair

What sinks the answer

  • "Reliable disks" only
  • No notion of redundancy/erasure coding
  • Unaware of corruption detection

If you genuinely do not know

Say this instead of freezing. Reasoning out loud from what you do know beats silence every single time, and a good interviewer is listening for exactly that.

High durability comes from [redundancy, not reliable disks]: [multiple copies across many devices/AZs so no single failure loses data], often [erasure coding, data + parity shards reconstructable from a subset, efficient redundancy]. Plus [checksums to detect/repair bit rot] and [self-healing, auto-rebuild missing copies when a device fails]. Eleven nines is [the product of all of these].

Keep going with storage

All 336 cloud engineer questions

Knowing the answer is not the same as recalling it under pressure

Sign in to send the questions you fumble to spaced recall, so they come back right before you would forget them, and learn the concepts behind them with hands-on labs.

Start free