- remove docker/docker from Traefik vendor (unused)
- use `ignore` for all Traefik vendor in integration glide.
- defined only integration specific version of the dependencies.
Log message produced by go-marathon was:
time="2017-06-28T09:08:19Z" level=debug msg="listenToSSE(): failed to
handle event: failed to decode the event type, content: , error: EOF"
The fix for this was done in the upstream project of go-marathon
donovanhide/eventsource.
Background is that Marathon periodically sends a \n over the SSE
subscription, in order to keep the connection alive. This was parsed as
empty event by the eventsource and published. go-marathon in turn was
not able to do something with this empty event was producing the log
message above. By getting rid of publishing empty events in the
downstream library, we also get rid of this log message.
Fixes a problem with UnreachableStrategy being available now in two
type-incompatible formats (object and string).
We also upgrade the transitive dependency
github.com/donovanhide/eventsource.
In the event that a user needs to explode their acme.json file into
a set of directories and relevant files for troubleshooting or use
with other programs this script will parse them into the components
in the following path structure:
```
certdir
├── certs
│ ├── domain-1.example.com
│ ├── domain-2.example.com
│ └── domain-n.example.com
└── private
└── letsencrypt.key
```
Introduces Rancher's metadata service as an optional provider source for
Traefik, enabled by setting `rancher.MetadataService`.
The provider uses a long polling technique to watch the metadata service and
obtain near instantaneous updates. Alternatively it can be configured to poll
the metadata service every `rancher.RefreshSeconds` by setting
`rancher.MetadataPoll`.
The refactor splits API and metadata service code into separate source
files respectively, and specific configuration is deferred to
sub-structs.
Incorporates bugfix #1414
Now retries only happen when actual network errors occur and not only
anymore based on the HTTP status code. This is because the backend could
also send this status codes as their normal interface and in that case
we don't want to retry.
