-
Notifications
You must be signed in to change notification settings - Fork 621
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
vendor: update gRPC, protobuf #2452
Conversation
Looks like I have some work to do;
|
Due to this change; grpc/grpc-go@596a6ac#diff-b3a9f1c1b5d5600f4924efe942b13fb0 updating now |
c130a68
to
23df0f5
Compare
LGTM when the tests pass. Let me know if you need help hunting down the correct context functions. |
I replaced for
I'll do some digging to see where the context is changed, so the other one should be used 👍 |
23df0f5
to
d94498c
Compare
Codecov Report
@@ Coverage Diff @@
## master #2452 +/- ##
==========================================
- Coverage 61.78% 61.76% -0.03%
==========================================
Files 134 134
Lines 21829 21821 -8
==========================================
- Hits 13488 13478 -10
- Misses 6887 6899 +12
+ Partials 1454 1444 -10 |
@thaJeztah are you still working on this? |
I was looking this afternoon to see if I could find where changes are needed, but didn't spot it directly 😓 Current failure is:
|
if rb == nil { | ||
// TODO(bar) return error when DNS becomes the default (implemented and | ||
// registered by DNS package). | ||
grpclog.Infof("could not get resolver for scheme: %q", target.Scheme) |
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.
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.
Not sure why we're having these issues lately. I've restarted the CI.
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.
LOL, now it's even worse; definitely possible it's related to this PR, but don't have a clue yet where the problem would be
8c97b32
to
079fc3c
Compare
4f9ecee
to
ee0bb7a
Compare
Can we get some maintainers looking at this problem? We are trying to fix this but there are some larger issues that swarm it panicing over version bumps. |
We need to get moby/moby#36895 into the next release, what can we do to help get this unblocked? |
We'll do this on priority. Restarting CI one more time to judge if the failure is due to flakiness, but will look into the diff to try and figure out. |
@nishanttotla it needs to be rebased, just in case. |
ee0bb7a
to
e788cf4
Compare
rebased |
@thaJeztah it seems like you'll have to remove |
Hm, let me try |
e3b3cb8
to
2a14d4a
Compare
s, _ := status.FromError(err) | ||
assert.Equal(t, codes.Internal, s.Code()) | ||
assert.Equal(t, codes.ResourceExhausted, s.Code()) |
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.
Is this change needed along with the other changes ?
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.
I think so (it was a while back that I opened this PR);I think they changed these errors
i'm pouring through the logs on these test failures and i'm pretty confused. it seems to me that something may have changed in the internals of grpc, but i can't figure out what it would have been. |
I did notice there are more recent versions since I created this PR; would it be worth updating to a more recent version? |
I'm unsure. It's certainly worth trying, to see if it will quietly start passing again. I'm digging into the root cause right now. |
agreed. it makes sense to try upgrading to the latest version @thaJeztah |
I figured it out, 1 sec, verifying the fix. |
@thaJeztah I pushed to your branch a fix, I hope you don't mind. The problem was that we were string matching an error message, but the message changed. |
3ed989f
to
8dc087b
Compare
} | ||
|
||
// additionally, the error we're looking for is an internal error | ||
if grpcStatus.Code() != codes.Internal { |
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.
Wondering; does it return a specific .Code()
that could be useful for detection (in addition to string-matching)?
// MaxMsgSize returns a ServerOption to set the max message size in bytes for inbound mesages. | ||
// If this is not set, gRPC uses the default 4MB. | ||
// MaxMsgSize returns a ServerOption to set the max message size in bytes the server can receive. | ||
// If this is not set, gRPC uses the default limit. Deprecated: use MaxRecvMsgSize instead. | ||
func MaxMsgSize(m int) ServerOption { |
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.
@dperny Just saw this was marked deprecated; perhaps you can update https://github.com/docker/swarmkit/blob/8aa9c33bcdff9ea38fc79e0b1d054199917513f3/manager/manager.go#L272, which is using the deprecated function
t.Close() | ||
return nil, connectionErrorf(true, err, "transport: %v", err) | ||
return nil, connectionErrorf(true, err, "transport: failed to write window update: %v", err) |
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.
Looks indeed that ConnectionError
doesn't have any additional information 😞
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.
LGTM, but a squash is needed.
commits looked sensibly grouped, but not a maintainer |
I think 4, 5 should be squashed into 1. I don't think these commits mean much individually. |
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Fixes an issue resulting from string matching the error result of grpc methods. Adds comments documenting the fix, as well as explaining how to avoid this issue in the future. Also updates manager/manager.go to use the function MaxRecvMsgSize, instead of the deprecated equivalent MaxMsgSize. Signed-off-by: Drew Erny <drew.erny@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
3cac027
to
8a2b6fd
Compare
rebased and squashed |
Oh, I may have mis-interpreted; you want the changes into the same commit as the revendor? (in moby/moby we usually keep local changes separate from the vendor commit) |
I guess that makes sense. Thanks for squashing the last two! |
This updates the vendored packages for gRPC, protobuf, and re-generates (
make generates
)Full diffs of the updated dependencies:
NOTE: this PR is built on top of #2451, so needs to be rebased if that one is merged; I thought keeping them separate was cleaner, but we can close the other PR if combining them is okay 😄
relates to #2448, #2444
ping @stevvooe @aaronlehmann @dperny @anshulpundir PTAL