Hot, warm and cold describe storage tiers for security data, ranked by how quickly the data can be searched and how much it costs to keep. Hot data is recent, fully indexed and fast to search, and costs the most. Warm data is older and cheaper, and usually still searchable, though sometimes more slowly. Cold data is archived at the lowest cost and often has to be restored before anyone can search it.
Hot searchable data is the data analysts and detections can query right away, without waiting or filing a request.
The exact boundaries vary by organization, tool and regulation.
The problem is that cold often means unusable in practice. If searching data older than 90 days takes a restore request and a day of waiting, analysts rarely do it, even when an investigation needs it. Attacks discovered late, which is common, then cannot be fully traced. Many organizations end up paying to keep data they effectively never look at.
Newer security data lake designs blur the tiers by keeping older data in low-cost storage that can still be queried directly. See fast search across full-fidelity data.
Is cold storage searchable?
Sometimes. Traditional cold archives usually require restoring data first. Some modern security data lakes keep low-cost storage directly searchable.
How much security data should be hot?
Enough to support real-time detection and most active investigations, often the last few weeks. The more important question is how much older data stays directly searchable.
The Anomali Unified Security Data Lake is designed to keep full-fidelity telemetry searchable for long periods at low cost, so older data stays useful for investigations.
Read Full fidelity at cold storage cost or request a demo.