This is an attempt to fix the flaky test of TestSwarmNodeTaskListFilter in 25029.
Basically this fix adds a check to wait until 3 containers has already up,
before processing `node tasks ...`.
This might fix 25029.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
Upstream-commit: 63c0366bc9a7f326ad61cdfa69fa860cf844a2c7
Component: engine
This fix tries to address the issue raised in 24881 where
the stack name is not visable when inspecting containers
deployed through stacks.
This fix adds `labelNamespace` (`com.docker.stack.namespace`)
to containers at the client side so that the stack name label
will show up after containers are deployed.
This fix is tested manually:
- Build the binary with `make DOCKER_EXPERIMENTAL=1`
- Create a bundle file and deploy with the `docker stack deploy`
- Verify that `com.docker.stack.namespace` has been
set properly by checking contaners with `docker inspect <ContainerID>`
This fix fixes 24881.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
Upstream-commit: 796636716d547dc7c37ff98950ba2c662143e555
Component: engine
In `TestApiSwarmRestartCluster`, it's calling `checkClusterHealth`.
`checkClusterHealth` calls `d.info()`, which will return an error if
there is no cluster leader... problem is `checkClusterHealth` is doing a
nil error assertion w/o giving any time for a leader to be elected.
This moves the `d.info()` call into a `waitAndAssert` using the default
reconciliation timeout.
It also moves some other checks into a `waitAndAssert` to give the
cluster enough time to come back up.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Upstream-commit: fdcde8bb65acf2c459e93678bf2659139ef5e918
Component: engine
Signed-off-by: John Howard <jhoward@microsoft.com>
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
Upstream-commit: 755be795b4e48b3eadcdf1427bf9731b0e97bed1
Component: engine
`TestParseWords` needs to use the `tokenEscape` for one of the test
cases, but `tokenEscape` was not being set unless tests ran in a
specific order.
This sets a default value for `tokenEscape`... `\`... so that tests that
rely on this global are not affected by test ordering.
This is the simplest fix for these cases. Ideally the token should not
be set as a global but rather passed down, which is a much larger
change.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Upstream-commit: df167d3ff04cdc90012c8ca39647662ad69e6715
Component: engine
With digests being added by default, all images have multiple references.
The check for whether force is required to remove the reference should use the new check for single reference which accounts for digest references.
This change restores pre-1.12 behavior and ensures images are not accidentally left dangling while a container is running.
Signed-off-by: Derek McGowan <derek@mcgstyle.net> (github: dmcgowan)
Upstream-commit: 1f7a9b1ab3d261de5be7d490e7e4f978f317242f
Component: engine
When updating values in the spec according to CLI flags, don't write
into the existing pointers. They may be nil. Instead, update them to
point to the new value we're writing.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
Upstream-commit: f9c920a1266197c2b6f0378b58f22246037fda7a
Component: engine
Context cancellations were previously causing `Prepare` to fail
completely on re-entrant calls. To prevent this, we filtered out cancels
and deadline errors. While this allowed the service to proceed without
errors, it had the possibility of interrupting long pulls, causing the
pull to happen twice.
This PR forks the context of the pull to match the lifetime of
`Controller`, ensuring that for each task, the pull is only performed
once. It also ensures that multiple calls to `Prepare` are re-entrant,
ensuring that the pull resumes from its original position.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
Upstream-commit: d8d71ad5b94d44a2778f2d8989424259cac94e9b
Component: engine
Remove the swarm inspect command and use docker info instead to display
swarm information if the current node is a manager.
Signed-off-by: Vincent Demeester <vincent@sbr.pm>
Upstream-commit: e6923f6d75c2bd1b22cc1229214ffceca3251cc6
Component: engine
When using encrypted vxlan network, some of the xfrm states are left
stale. This fix also filters out self advertise-addr rules.
Signed-off-by: Madhu Venugopal <madhu@docker.com>
Upstream-commit: 6a754a4eef4463ed7c712a1ebe5bc38e2a957c42
Component: engine
This prevents unnecessary API call to containerd.
Signed-off-by: Anusha Ragunathan <anusha@docker.com>
Upstream-commit: b867f6c6e18f9d999ddec911a241941428777569
Component: engine
Swarm mode makes it possible through the API to set labels to containers
but not through command line. This tries to fix it.
Signed-off-by: Vincent Demeester <vincent@sbr.pm>
Upstream-commit: 4031d70d1be2f02e1c9d6f7738f2a816be7eeef2
Component: engine
Ensure that cancellation of a pull propagates rather than continuing to
container creation. This ensures that the `Prepare` method is properly
re-entrant.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
Upstream-commit: d99c6b837ffd18ffe5bce801feb4936bf0edd2aa
Component: engine
Yong Tang
Brian Goff
Charles Smith
mapk0y
Anusha Ragunathan
John Howard
Vincent Demeester
Sebastiaan van Stijn
lixiaobing10051267
Harald Albers
Tibor Vass
Akihiro Suda
Derek McGowan
Aaron Lehmann
Tibor Vass
Stephen J Day
Madhu Venugopal
Nishant Totla
runshenzhu