LindormTSDB stores time series data across three tiers — hot, warm, and cold — and automatically moves aging data to Capacity storage to reduce costs. This document explains how to enable cold storage and configure the archiving boundary.
How it works
As time series data ages, it is queried less frequently. LindormTSDB separates data into three tiers based on recency and query frequency:
| Tier | Characteristics |
|---|---|
| Hot data | Most recently written; timestamps close to the current time; queried frequently |
| Warm data | Stored for some time; queried less frequently than hot data |
| Cold data | Stored for a long time; rarely queried; accounts for the largest share of storage |
The following diagram shows how LindormTSDB implements tiered storage.
Prerequisites
Before you begin, ensure that you have:
A Lindorm instance with LindormTSDB activated
Sufficient Cold Storage Capacity to accommodate your archived data
Enable cold storage
By default, a Lindorm instance with LindormTSDB provides tiered storage for hot and warm data only. To archive cold data to Capacity storage automatically, enable cold storage explicitly.
Enable cold storage using one of the following methods:
At instance creation: On the buy page, configure the Purchase Cold Storage and Cold Storage Capacity parameters.

After instance creation: In the Lindorm console, navigate to the instance settings and enable cold storage there.

After enabling cold storage, set the hot and cold data boundary to control when data is archived.
Set the hot and cold data boundary
What the boundary means
The hot and cold data boundary is a threshold, measured in days, that determines when a data record is treated as cold and moved to Capacity storage. LindormTSDB applies the following formula to each data record:
Current timestamp - Timestamp of a data record >= Hot and cold data boundary (days)LindormTSDB archives data automatically once the threshold is met — no manual intervention is required.
Configure the boundary
Set the hot and cold data boundary using either of these methods:
SQL statement: Specify the boundary in a
CREATE DATABASEorALTER DATABASEstatement. For syntax details, see CREATE DATABASE or ALTER DATABASE.Lindorm console: Set the boundary in the database settings page. For steps, see Manage databases.
Query cold data
Query cold data using the same standard SQL statements you use for hot or warm data — no additional conditions or configuration changes are needed.
Queries against cold data archived in Capacity storage take longer to return results than queries against hot or warm data. Capacity storage is optimized for cost, not read latency. If low-latency access to a dataset becomes a priority, adjust the boundary so that data remains in the hot or warm tier longer.