-
Notifications
You must be signed in to change notification settings - Fork 40.1k
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
More corner cases with iptables proxy #12764
Conversation
Read through current last commit (thockin@23a3ab8). |
GCE e2e build/test passed for commit 23a3ab8355eb48317dc2860a394ba76b5001a473. |
be9a28c
to
c4e2c83
Compare
GCE e2e build/test passed for commit be9a28ca34ebf46c76fcb6000f96684dfdb4332a. |
GCE e2e build/test passed for commit c4e2c83f304966d7042819cd80ab512104856912. |
GCE e2e build/test passed for commit 392c340c852747e6f372ab9d82aeeb10e1ba4998. |
392c340
to
35dd764
Compare
GCE e2e build/test passed for commit 35dd764ce77dbcbc0be77f74775c00bbf0770588. |
35dd764
to
e93a734
Compare
GCE e2e build/test passed for commit e93a7340c74cef99f7d78c1c5e5d45e84568ca72. |
e93a734
to
b55b1b4
Compare
GCE e2e build/test passed for commit b55b1b400d34582c0ccf76cad969c4cccad7f1dd. |
args = append(args, | ||
"-m", "statistic", | ||
"--mode", "random", | ||
"--probability", fmt.Sprintf("%f", 1.0/float64(n-i))) |
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.
(discussed offline)
we should probably restrict decimal length
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.
Done in next push
On Mon, Aug 17, 2015 at 4:30 PM, Abhi Shah [email protected] wrote:
In pkg/proxy/iptables/proxier.go
#12764 (comment)
:}
if i < (n - 1) {
// Each rule is a probabilistic match.
args = append(args,
"-m", "statistic",
"--mode", "random",
"--probability", fmt.Sprintf("%f", 1.0/float64(n-i)))
(discussed offline)
we should probably restrict decimal length—
Reply to this email directly or view it on GitHub
https://github.com/kubernetes/kubernetes/pull/12764/files#r37249463.
b55b1b4
to
6f34be3
Compare
Issues fixed, new push is up. This needs e2e before it can be activated but I'd like to merge it ASAP. |
GCE e2e build/test failed for commit 6f34be3. |
@k8s-bot test this please |
GCE e2e build/test passed for commit 6f34be3. |
LGTM |
More corner cases with iptables proxy
Whoo!
|
So we still need hairpin mode, anything else? |
hairpin mode (hairy) Not sure what else. If you have extra bandwidth, you could do the "always On Tue, Aug 18, 2015 at 9:18 AM, Benjamin Elder [email protected]
|
I'll take a look at the always masq flag this afternoon. As far as I know On Tue, Aug 18, 2015 at 12:25 PM, Tim Hockin [email protected]
|
yeah, maybe. It's P2 or P3 to me. The other thing we could use is better On Tue, Aug 18, 2015 at 9:32 AM, Benjamin Elder [email protected]
|
Yeah, I'll see if I can get a bit more done early this week before i get On Tue, Aug 18, 2015 at 12:36 PM, Tim Hockin [email protected]
|
Node ports
Set sysctls
Cleanup rules code
Faster sync at startup
Better logging
DNAT back to self doesn't quite work yet because we need to enable hairpin mode on the bridge, and that has to be done interface-by-interface (sigh). Will keep hunting for a better answer there.
This needs SERIOUS e2e testing. I have a very manual battery of tests I have been running, which is a good starting place.
@BenTheElder @ArtfulCoder