Skip to content

Update logrus to v1.0.1 (Sirupsen -> sirupsen)#34272

Merged
lowenna merged 2 commits intomoby:masterfrom
dmcgowan:update-logrus
Aug 2, 2017
Merged

Update logrus to v1.0.1 (Sirupsen -> sirupsen)#34272
lowenna merged 2 commits intomoby:masterfrom
dmcgowan:update-logrus

Conversation

@dmcgowan
Copy link
Member

@dmcgowan dmcgowan commented Jul 26, 2017

Alternative to #33998

This handle the case sensitivity issue with logrus changing names. This is currently using manual edits on some of the changes in the vendor directory until the vendored projects merge their associated changes. This change aims to have minimal impact in terms of new vendored code. Each of the individual libraries should have changes after the updates get merged.

See https://github.com/sirupsen/logrus#logrus- for the reasoning behind this change

Related PRs

@AkihiroSuda AkihiroSuda changed the title Update logrus to v1.0.1 Update logrus to v1.0.1 (Sirupsen -> sirupsen) Jul 27, 2017
@AkihiroSuda
Copy link
Member

I modified the title of this PR from Update logrus to v1.0.1 to Update logrus to v1.0.1 (Sirupsen -> sirupsen) so that 3rd parties importing client pkg can easily find this PR, but please feel free to revert 😄

Copy link
Member

@dnephin dnephin left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@aaronlehmann
Copy link

I'm confused about how this passes CI. I'd expect the "dirty" vendor directory to fail the vendor validation check, and I don't see changes to the validation scripts to make an exception.

@dnephin
Copy link
Member

dnephin commented Jul 27, 2017

hmm, good question. It looks like the vendor job didn't run on this.

@dmcgowan
Copy link
Member Author

Updated to make the vendor done more cleanly. It is using forks where the upstreams have not been merged. The forks are based on the version which was previously there with only a commit to rename the imports. When the upstreams are merged and go to update the corresponding vendor, then updated imports should work fine. If any of these dependencies try to update before merging the logrus change upstream (they are all docker-related projects), then they will have to fix logrus themselves or merge the change (they will not have my sympathy).

@dnephin
Copy link
Member

dnephin commented Jul 28, 2017

Any chance your branch can be opened as a PR against those repos ? If it doesn't merge cleanly maybe just an issue to track the change?

@lowenna
Copy link
Member

lowenna commented Aug 1, 2017

Restart z again as the previous run had hit at least one failure:

18:26:25 ----------------------------------------------------------------------
18:26:25 FAIL: docker_cli_daemon_test.go:1416: DockerDaemonSuite.TestCleanupMountsAfterDaemonAndContainerKill
18:26:25 
18:26:25 [d81792b14d8fa] waiting for daemon to start
18:26:25 [d81792b14d8fa] daemon started
18:26:25 
18:26:25 docker_cli_daemon_test.go:1437:
18:26:25 [d81792b14d8fa] exiting daemon
18:26:25     // restart daemon.
18:26:25     d.Restart(c)
18:26:25 daemon/daemon.go:392:
18:26:25     t.Fatalf("Error while stopping the daemon %s : %v", d.id, err)
18:26:25 ... Error: Error while stopping the daemon d81792b14d8fa : signal: killed
18:26:25 
18:26:29 
18:26:29 ----------------------------------------------------------------------

Copy link
Member

@vdemeester vdemeester left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dmcgowan let's update the vendor.conf to not point on forks for opengcs, runc and containerd (with containerd/containerd#1252 merged) and use forks for libnetwork and swarmkit to deal with the chicken-and-egg problem

github.com/gorilla/context v1.1
github.com/gorilla/mux v1.1
github.com/jhowardmsft/opengcs v0.0.9
github.com/jhowardmsft/opengcs b9d0120d36f26e981a50bf18bac1bb3f0c2b8fef https://github.com/dmcgowan/opengcs.git
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dmcgowan PR has been merge, we should update to not using the fork anymore 👼

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Merged but cannot use master yet, see conversation with @jhowardmsft about a follow up to bring it to master. I am using a fork here to avoid pulling in unrelated changes of which I know little about. My expectation is those that better understand these projects will be able to get off my forks by opening a follow up updating to the most recent.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@vdemeester Agreed with @dmcgowan's comment ^^. I have exactly that follow-up PR I'm about to submit which will be backed up behind this one being merged, and move back to jhowardmsft/opengcs, but at v0.0.12, which incorporates lower-case sirupsen

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ok sounds fair 👼


# When updating, also update RUNC_COMMIT in hack/dockerfile/binaries-commits accordingly
github.com/opencontainers/runc 2d41c047c83e09a6d61d464906feb2a2f3c52aa4 https://github.com/docker/runc
github.com/opencontainers/runc e9325d442f5979c4f79bfa9e09bdf7abb74ba03b https://github.com/dmcgowan/runc.git
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here 👼

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here about follow ups, I have already discussed with @crosbymichael about getting us back onto github.com/opencontainers/runc after this change.


# containerd
github.com/containerd/containerd 3addd840653146c90a254301d6c3a663c7fd6429
github.com/containerd/containerd fc10004571bb9b26695ccbf2dd4a83213f60b93e https://github.com/dmcgowan/containerd.git
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should wait for containerd/containerd#1252 to get merged before merging this one 👼

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's merged and can be updated here 👍

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same, this should be done in follow up so the containerd related changes can be reviewed separately. My fork only applies the logrus changes on top of the previous hash

@lowenna
Copy link
Member

lowenna commented Aug 1, 2017

Restarted z CI (yet again...)

@dmcgowan
Copy link
Member Author

dmcgowan commented Aug 1, 2017

@vdemeester I probably should have outlined the plan here better. The points of these forks is intended to be temporary since updating the dependencies will require removing my fork to pick up any functionality changes. I have tried to link to all the other work not as blockers to get this in, but to show the follow up work is in progress to get us back to the ideal fork-less state. Unfortunately this is a much messier change anyone would have liked and merging it will help us move past it quicker 😄

@vdemeester
Copy link
Member

@dmcgowan sounds good 👍 Let's merge this 👼

Copy link
Member

@johnstep johnstep left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@thaJeztah
Copy link
Member

Thanks for explaining @dmcgowan - @vdemeester and I were discussing this morning, and thought it was best to first have the upstreams merged, so that we would be sure it would be accepted upstream; LGTM with that explanation 👍

@lowenna
Copy link
Member

lowenna commented Aug 1, 2017

And restarted z again. Failed almost 3 hours in. @thaJeztah Who can whack the z nodes into submission? 7, for example, seems to be running under half the speed of node 1.

(Latest iteration is running on node 1 which seems to be one of the most reliable of the current set of nodes)

@lowenna lowenna merged commit 8af4db6 into moby:master Aug 2, 2017
@thaJeztah
Copy link
Member

ping @tophj-ibm ^^

@tonistiigi
Copy link
Member

@dmcgowan Is there an issue tracking the docker/cli update. I don't see it referenced here.

@vdemeester
Copy link
Member

@tonistiigi @simonferquel created a PR for it 👼

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.