Issue metadata
Sign in to add a comment
|
Lack of documentation around FCM throttling
Reported by
bhhro...@justin.tv,
Jun 13 2017
|
||||||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.86 Safari/537.36 Steps to reproduce the problem: 1. Create a push subscription. 2. Send a message using that push subscription. Receive a 201 and view push message being rendered. 3. Send the same message using that push subscription. Receive a 201, but notice that message does not render. 4. Change message in some way and resend. Sometimes the browser notification will render, sometimes not. There seems to be a minimum mandatory difference in message payloads as well as time between messages in order for the message to be successfully sent from the push service. What is the expected behavior? There appears to be throttling of some sort going on, however, I have been unable to find documentation of any sort around this. While I understand that throttling is certainly necessary to combat spam, prevent DoS attacks, etc, it would be nice to have some more documentation around throttling criteria as well as more explicit status codes. What went wrong? Apparent throttling of FCM push service, without any signal from status codes or mention within documentation. Did this work before? N/A Does this work in other browsers? N/A Chrome version: 59.0.3071.86 Channel: stable OS Version: OS X 10.11.6 Flash Version: Shockwave Flash 26.0 r0
,
Jun 14 2017
@bhhronik-- Could you please provide us the sample test file/url to reproduce the issue and also please help us by providing the screenshot or screencast of the expected result ,that would really helpful to traige issue better. Thanks!
,
Jun 19 2017
,
Jun 20 2017
Did you observe on chrome://gcm-internals/ that the message indeed wasn't received by Chrome? It's rather hard to trigger FCM throttling right now, so I think we have to look at a number of other options first. Did you observe this just on Mac? We switched to the Mac system notification center in Chrome 59, which removes a bunch of cases where we'd have throttled the notifications (for instance because another app was full screen).
,
Jun 24 2017
Hi Peter, I have opened another issue, which may be related: https://bugs.chromium.org/p/chromium/issues/detail?id=736442. In there I have provided screenshots of all of the gcm internals. After inspecting the gcm-internals, I am noticing that the browser is receiving the push message fine and that there is no decryption error, however it does not appear to be triggering the push event handler of the registered service worker. Unfortunately, I did not inspect the gcm-internals upon upon up the ticket around FCM throttling, so I cannot say for certain whether or not they are the same issue.
,
Jun 24 2017
Thank you for providing more feedback. Adding requester "hdodda@chromium.org" to the cc list and removing "Needs-Feedback" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jun 26 2017
Let's continue discussion on Issue 736442 then. Especially since you haven't launched yet (to my knowledge) it's extremely unlikely that you've hit any sort of server-side throttling systems. If you do suspect this happens in the future, it'd be grand if you could share the subscription information (as you did in some of the other issues) as it enables us to investigate further. |
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by bhhro...@justin.tv
, Jun 13 2017