Skip to content
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

DDFS: enable OS controlled server socket size #613

Open
wants to merge 1 commit into
base: develop
Choose a base branch
from

Conversation

jgrnt
Copy link

@jgrnt jgrnt commented Mar 16, 2015

Fixes #603
Sets the TCP socket buffer size to OS controlled

Based on changes in current mochiweb upstream mochi/mochiweb#153, therefore update mochiweb version.
Streaming body size set to the previous default socket size(8KB).
Should be further investigated if the size influences performance or should be configurable.
Did a small test with 1MB, but did not change results.

Based on changes in current mochiweb upstream
Streaming body size set to the previous default socket size(8KB).
Should be further investigated if the size influences performance or should be configurable.
@erikdubbelboer
Copy link
Contributor

@pooya Any idea when this will be merged? Are you still maintaining disco or is anyone else maintaining it?

@pooya
Copy link
Member

pooya commented Jun 16, 2015

@erikdubbelboer I have to test this on a multi-node cluster to make sure it does not break anything. Currently, I have limited resources for testing the patches (I have to spin up a cluster on GCE).

@erikdubbelboer
Copy link
Contributor

I see you're not working at chango anymore so I guess working on Disco isn't part of your job anymore? Let me know if there is any way I can help.

@pooya
Copy link
Member

pooya commented Jun 20, 2015

Well, it was not part of my job in chango either, but I had a lot of free time :) The Disco team has left and joined Mozilla and they are still using / developing Disco.
You can always help by testing the patches. Mainly, if https://github.com/pooya/discostuff/blob/master/bisect_script.sh passes with a patch in a cluster of three nodes or larger, I have no objection for merging the patch.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Slow uploads with high latency
3 participants