---
id: whisper
version: "1.1.10"
license: Apache Software License 2.0
license_treatment: permissive
maintenance: abandoned
---

# whisper — Fixed size round-robin style database

License: permissive · Maintenance: abandoned · Downloads: 547.7K/mo

## What it is and what it does

Whisper is a standalone time-series database library designed to store numeric metrics efficiently in a fixed-size file format, similar to RRD (round-robin-database). It is part of the Graphite monitoring ecosystem but can be used independently. The library supports multiple archive definitions, allowing recent data to be stored at high resolution while older data automatically rolls up into lower-resolution archives for long-term retention without unbounded disk growth.

The package provides a command-line toolkit for database operations: creating whisper files with custom retention policies, updating them with timestamped values, fetching and dumping data, resizing archives, merging files, and inspecting metadata. It has no external runtime dependencies, making it lightweight to integrate into monitoring pipelines or custom metric collection systems. However, the project is abandoned with the last release on 2022-05-22, meaning no active maintenance, security updates, or compatibility work for modern Python versions.

Use it for:

- Store application or infrastructure metrics in a fixed-size database for a monitoring system without external dependencies.
- Migrate existing RRD-based metric storage to Whisper format using the rrd2whisper.py conversion tool.
- Build a custom time-series backend that degrades data resolution over time to control storage footprint.
- Query and export historical metrics from Whisper files for analysis or reporting via whisper-fetch.py.
- Merge or fill gaps in metric data across multiple Whisper files during data recovery or consolidation.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Whisper is a fixed-size time-series database library that stores numeric data at multiple resolutions, degrading older data from high to low resolution for long-term retention.

No, not recommended for new projects. While Whisper is a stable, dependency-free time-series database, it is abandoned with no maintenance since 2022-05-22 and no active development. Existing Graphite installations may continue to use it, but new deployments should consider maintained alternatives. Security issues and Python version incompatibilities are unlikely to be addressed.

## Install

pip install whisper
uv add whisper
poetry add whisper

## Installing whisper

Before you install: Installation friction is high. The package is abandoned with no recent maintenance, though it remains in the top 15000 by downloads. No runtime dependencies simplifies installation, but the lack of active maintenance means security updates and bug fixes are unlikely.

License in practice: Licensed under Apache Software License 2.0, a permissive license that allows commercial and private use with minimal restrictions, making it safe for most projects from a licensing standpoint.

Quickstart:

pip install whisper

from whisper import create, update, fetch

create('/path/to/metric.wsp', [(60, 1440)])
update('/path/to/metric.wsp', value, timestamp)
from_time, until_time, step, values = fetch('/path/to/metric.wsp', from_time)

No external runtime dependencies, but the package is abandoned and may have compatibility issues with Python versions beyond those listed in classifiers (3.7).

Verify before relying:

- Whether the package works reliably on Python versions beyond 3.7 despite classifier claims.
- Current state of compatibility with modern operating systems and file systems.
- Whether the Graphite project has superseded or replaced Whisper with a maintained alternative.

## Package facts

- License: Apache Software License 2.0 (permissive)
- Python support: unspecified
- Install friction: high
- Maintenance: abandoned
- Downloads: 547.7K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

time series database, round robin database, metric storage, whisper database format, numeric data retention, fixed size database, time-series backend, time-series-storage, round-robin, abandoned

[View on SkillFed](https://skillfed.io/packages/whisper) · [View on PyPI](https://pypi.org/project/whisper/)