# Bufferbloat Test

Measure how much latency increases when your connection is busy downloading and uploading, and what that added delay can mean for calls and video.

## Ready to Test

Measure idle latency, then latency under download and upload load.

HTTP round trips to the same ConnectionCheck service as our ping test. Retransmission, browser scheduling and server processing can affect these measurements.

Enable JavaScript to run the test.

## What This Test Measures

Bufferbloat is extra delay when traffic queues build up under load. This test compares median HTTP round-trip latency at idle with medians during separate download and upload phases. Each load phase lasts up to eight seconds, using two transfers at once on a separate server port. It reports the larger increase, treating a negative increase as zero for grading.

The byte and time limits can prevent a fast connection from being fully loaded. The server, browser, Wi-Fi or route can also limit transfers. Results describe the load achieved, not proof that the broadband link was saturated. An incomplete phase receives no grade. This is not a packet-loss test or a certified speed measurement.

## How We Grade the Increase

These are this test's diagnostic thresholds, not an industry standard or a call-quality guarantee. The basis is the extra round-trip delay observed here: the bands separate tens of milliseconds from a tenth of a second or more. They are a local interpretation aid, not measured application failure points.

- **Low Increase:** up to 20 ms. Little added delay was observed at this load.
- **Moderate Increase:** over 20 to 50 ms. Calls and interactive video may become less responsive under load.
- **High Increase:** over 50 to 100 ms. Extra delay may make conversation harder when transfers run.
- **Very High Increase:** over 100 ms. Substantial added delay was observed; investigate congestion before relying on busy-period calls.

Even a low increase can sit on top of high idle latency. Calls also depend on jitter, loss and the route to the calling service. Run the [ping, jitter and packet loss test](https://netavo.com/tools/ping-test) for that separate view. Compare a wired test with Wi-Fi and repeat when the site is quiet. Router queue management and scheduling large transfers are areas to investigate.

For capacity planning, use the [broadband speed calculator](https://netavo.com/tools/broadband-speed-calculator).

## Frequently Asked Questions

### Does a Low Increase Guarantee Good Calls?

No. Netavo reports the increase at the load achieved. Idle latency, jitter, packet loss and the route to the calling service also affect calls.

### Why Can the Result Be Inconclusive?

The Netavo test has byte and time limits. A blocked endpoint, failed transfer, interrupted tab or insufficient measurements prevents a complete comparison and receives no grade.

## Connectivity for Your Business

Discuss your site's download, upload and call requirements with Netavo.

[Business Broadband](https://netavo.com/services/business-broadband)

[Work with us](https://netavo.com/work-with-us)

Source: https://netavo.com/tools/bufferbloat-test