-
Notifications
You must be signed in to change notification settings - Fork 203
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
Reduce network IOs by amplifying prefetch requests #609
Conversation
@jiangliu , a new test job has been submitted. Please wait in patience. The test job url: https://tone.openanolis.cn/ws/nrh4nnio/test_result/15806 |
@jiangliu , The CI test is completed, please check result:
Congratulations, your test job passed! |
size, | ||
state.clone(), | ||
)); | ||
rt.spawn(async move { |
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.
Why delegate the fetching data Future into the same tokio runtime rather than directly run to the end of future handle_prefetch_requests
. Moreover, it seems the runtime rt
only has a thread, so re-generate a future is efficent ?
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.
The current future acts as a dispatcher, which receives requests from the mpmc queue and dispatch them.
So there may be multiple inflight requests.
Currently two inlfight-requests are allowed.
Try to amplify prefetch request for v6 image to reduce network requests. Signed-off-by: Jiang Liu <gerry@linux.alibaba.com>
Limit number of inflight prefetch requests per thread, to give higher priority to on-demand requests. Signed-off-by: Jiang Liu <gerry@linux.alibaba.com>
@jiangliu , the code has been updated, so a new test job has been submitted. Please wait in patience. The test job url: https://tone.openanolis.cn/ws/nrh4nnio/test_result/15959 |
@jiangliu , The CI test is completed, please check result:
Congratulations, your test job passed! |
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
Reduce backend network IOs by enabling amplification of v6 prefetch requests. For a java image, it could reduce 50% network requests by average.