-
Notifications
You must be signed in to change notification settings - Fork 3.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fetch KeepAlive Quota Tests #4788
Fetch KeepAlive Quota Tests #4788
Conversation
Initial add of quota tests for keepalive fetch requests.
Notifying @jdm and @youennf. (Learn how reviewing works.) These tests will be available on w3c-test.org shortly after they are approved by a repository collaborator. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
These look good, just two minor nits. Curious, did you test them against an implementation?
|
||
function CreateKeepAliveRequest(delay, bodySize) { | ||
// Create a body of the specified size that's filled with *'s | ||
var requestBody = Array(bodySize).fill('*').join(""); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe use "*".repeat(bodySize)
instead?
var trickleURL = "../resources/trickle.py?count=1&ms="; | ||
var standardDelay = 1000; | ||
|
||
// We should expect 64KB of rolling quota for any type of keep-alive request sent. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
KiB
Yeah, validated using Edge's implementation. Rolling keepalive quotas will be available for fetch w/keepalive and sendBeacon in an upcoming insider build. |
Either something is wrong with the build system or these tests are flaky somehow. Not sure what is going on. I asked @jgraham to look into it. |
Okay, I think the problem is that you're putting your tests in w3c:MicrosoftEdge rather than master. That is what causes Travis to apply hundreds of commits and essentially get lost. |
This is breaking on travis because you are trying to merge into the MicrosoftEdge branch rather than master. We can probably fix that in the bot, but why are you trying to do that? |
We discussed this at the F2F last week, but I realize we did not come to a conclusion. Our plan was that we would use the MicrosoftEdge branch as the initial PR location where we'd do a Microsoft code review in the public eye. Then we'd do a 2nd PR from there merging it into master. We are trying to save other reviewers time so they don't review a change from Microsoft unnecessarily. I may not have been clear when we were discussing this and maybe I misunderstood suggestions. If there is a better way to save everyone time and get the public review from MS done, I'm happy to implement it. |
I don't mind reviewing tests from Microsoft directly. I also review them from Google and Apple, and sometimes even Mozilla. But I also don't mind if you want that kind of setup, but then the tooling should probably be updated to take that into account and not start pinging folks for pull requests that are not against master. |
Thanks Anne. I think for this PR, we will change it to just be against Master since the tools are not updated yet. Moving forward, we'll evaluate whether we should do some fork, modify the tools, or simply PR into Master and worry about code reviews there. |
I submitted #4878 against master for this change. @igrigorik @annevk |
Closing PR in favor of new one against master. |
Initial add of quota tests for keepalive fetch requests.