These tests don't validate if options are valid for the storage-driver,
nor do they test if the storage-driver itself is valid. However, the tests
were using actual values (such as s3) and options (such as "region") which
may lead to the conclusion that it's also testing validity of those values.
This patch replaces the test-values with non-existing driver-names and
options to make it more clear these are fake values.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Tiger Kaovilai <tkaovila@redhat.com>
This simple change mainly affects the distribution client. By respecting
the context the caller passes in, timeouts and cancellations will work
as expected. Also, transports which rely on the context (such as tracing
transports that retrieve a span from the context) will work properly.
Signed-off-by: Aaron Lehmann <alehmann@netflix.com>
The example shows a `GET` request to `/callback`, but it looks like it's
intending to show the `POST` request made by the registry to a
notificaitons endpoint. Unless I'm missing something, no `GET` request
should be involved.
Signed-off-by: Aaron Lehmann <alehmann@netflix.com>
The current doc says to use go get to pull the source code, which no
longer works by default on recent versions of go. Also, the makefile has
changed to not run all the validation automatically, and to use buildx
for some commands.
Signed-off-by: James Hewitt <james.hewitt@uk.ibm.com>
(*App).context, called in the HTTP handler on each request, creates a
URLBuilder, which involves calling Router(). This shows up in profiles a
hot spot because it involves compiling the regexps which define all the
routes. For efficiency, cache the router and return the same object each
time.
It appears to be safe to reuse the router because GetRoute is the only
method ever called on the returned router object.
Signed-off-by: Aaron Lehmann <alehmann@netflix.com>
Without this, the log message for the user indicates a problem with the
yaml file, so identifying the actual error is hard. This change fixes
the output so that the incorrect environment variable is easy to spot.
Fixes#3653
Signed-off-by: James Hewitt <james.hewitt@uk.ibm.com>
As the project has transitioned from Docker to CNCF, clean up the
contributing.md to reflect that:
- CNCF slack
- CNCF maintainers
- Use with other container tools
Fixes#3550
Signed-off-by: James Hewitt <james.hewitt@uk.ibm.com>
Current registry reference use a subset of dns and IPv4 addresses to
represent a registry domain.
Since registries are mostly compatible with rfc3986, that defines the
URI generic syntax, this adds support for IPv6 enclosed in squared
brackets based on the mentioned rfc.
The regexp is only expanded to match on IPv6 addreses enclosed between
square brackets, considering only regular IPv6 addresses represented
as compressed or uncompressed, excluding special IPv6 address
representations.
Signed-off-by: Antonio Ojea <antonio.ojea.garcia@gmail.com>
Go 1.16 reached end of life, so update to the current version of Go, but also
run CI on the previous version (which is still supported).
We should probably also decide wether or not we want the Dockerfiles to pin to
a specific minor version; this makes the releases more deterministic.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This reverts commit 06a098c632
This changes the function of linkedBlobStatter.Clear(). It was either removing the first of two possible manifest links or returning nil if none were found. Now it once again it removes only the valid manifest link or returns an error if none are found.
Signed-off-by: Bracken Dawson <abdawson@gmail.com>
