Closed Bug 1746796 Opened 3 years ago Closed 8 months ago

Websocket latency high gets worse over time

Categories

(Core :: Networking: WebSockets, defect, P2)

Firefox 95
defect

Tracking

()

RESOLVED INCOMPLETE

People

(Reporter: josroossien, Unassigned)

References

(Blocks 1 open bug, )

Details

(Keywords: parity-chrome, Whiteboard: [necko-triaged][necko-priority-queue])

Attachments

(1 file)

Attached image ping_firefox.png

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.110 Safari/537.36

Steps to reproduce:

I'm the developer of a realtime multiplayer browser game and people have been reporting a lot of lag lately on Firefox.

I created a small test script to test the ping and found out that the results are a lot worse on Firefox compared to Chrome when I run it side by side.
All I'm doing in the test is run a server on localhost which sends performance.now() to the client, the client will send this back to the server, the server calculates the difference and sends that to the client again so that the client can output it.

Here you can see the code I used for testing, I've attached a screenshot with the results.
https://github.com/Rojoss/websocket-ping-test

I ran the test on a clean install of Firefox in a private window, no extensions and such.
The websocket protocol version is rfc-6455 for both Chrome and Firefox.
Firefox version: 95.0.2 (64-bit)
Chrome version: Version 96.0.4664.110 (Official Build) (64-bit)

Actual results:

The ping on Firefox is super high and increases over time even when the window is focused. There are also way more spikes in the ping.

Expected results:

The ping should be similar to Chrome and not increase over time.

Component: Untriaged → Networking: WebSockets
Keywords: parity-chrome
Product: Firefox → Core
No longer blocks: websocket
Severity: -- → S3
Priority: -- → P2
Whiteboard: [necko-triaged]

Moving to priority queue for at least verification; if it does verify we'll want to understand this.

Thanks for the report and the testcase!! I'm sorry it fell through the cracks; we have a better system in place now for triaging bugs

Whiteboard: [necko-triaged] → [necko-triaged][necko-priority-queue]

I can't reproduce this with the provided test case.

Are you still able to reproduce this with the latest Firefox nightly version?
If yes, could you try to capture a http log? When capturing the log, could you also set New log modules: to timestamp,nsWebSocket:5?

Thanks.

Flags: needinfo?(josroossien)

Redirect a needinfo that is pending on an inactive user to the triage owner.
:kershaw, since the bug has recent activity, could you please find another way to get the information or close the bug as INCOMPLETE if it is not actionable?

For more information, please visit BugBot documentation.

Flags: needinfo?(josroossien) → needinfo?(kershaw)
Status: UNCONFIRMED → RESOLVED
Closed: 8 months ago
Flags: needinfo?(kershaw)
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: