-
Notifications
You must be signed in to change notification settings - Fork 16
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
Reuse the endpoint of the checkpointed container when restore #7
base: cr-combined
Are you sure you want to change the base?
Conversation
Methods for checkpointing and restoring containers were added to the native driver. The LXC driver returns an error message that these methods are not implemented yet. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: daemon/execdriver/native/create.go daemon/execdriver/native/driver.go daemon/execdriver/native/init.go
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Support was added to the daemon to use the Checkpoint and Restore methods of the native exec driver for checkpointing and restoring containers. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: api/server/server.go daemon/container.go daemon/daemon.go daemon/networkdriver/bridge/driver.go daemon/state.go vendor/src/github.com/docker/libnetwork/ipallocator/allocator.go
Restore failed if network resource not released during checkpoint, e.g., a container with port open with -p Signed-off-by: Hui Kang <[email protected]> Conflicts: daemon/container.go
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Add a basic test for checkpoint/restore to the integration tests Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Hui Kang <[email protected]>
@boucher This PR fixes the network endpoint problem when restoring a container. The endpoint should be reused by a checkpointed container so that the network resources (e.g., iptable rule) will not be changed after restoring. Please review if it fits into your branch. Thanks. |
If I run this code, and then checkpoint and restore a container, it still can't access the network. I believe this is related to the issue I filed here: moby#14027 (To test, I just exec'd into the restored container and tried ping'ing google.com) Are you actually able to use the network after a restore? |
Methods for checkpointing and restoring containers were added to the native driver. The LXC driver returns an error message that these methods are not implemented yet. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: daemon/execdriver/native/create.go daemon/execdriver/native/driver.go daemon/execdriver/native/init.go
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Support was added to the daemon to use the Checkpoint and Restore methods of the native exec driver for checkpointing and restoring containers. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: api/server/server.go daemon/container.go daemon/daemon.go daemon/networkdriver/bridge/driver.go daemon/state.go vendor/src/github.com/docker/libnetwork/ipallocator/allocator.go
Restore failed if network resource not released during checkpoint, e.g., a container with port open with -p Signed-off-by: Hui Kang <[email protected]> Conflicts: daemon/container.go
7bc31ef
to
af5a67b
Compare
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Add a basic test for checkpoint/restore to the integration tests Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Docker-DCO-1.1-Signed-off-by: Hui Kang <[email protected]> Merge branch 'cr-combined' into cr-combined-ep-fix Conflicts: daemon/container.go daemon/container_linux.go
@boucher Update the patch for 1.8.0-dev, but the problem you described still exists. I will look into this issue these days. |
Methods for checkpointing and restoring containers were added to the native driver. The LXC driver returns an error message that these methods are not implemented yet. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: daemon/execdriver/native/create.go daemon/execdriver/native/driver.go daemon/execdriver/native/init.go
Docker-DCO-1.1-Signed-off-by: Ross Boucher <[email protected]> (github: boucher)
Support was added to the daemon to use the Checkpoint and Restore methods of the native exec driver for checkpointing and restoring containers. Signed-off-by: Saied Kazemi <[email protected]> Conflicts: api/server/server.go daemon/container.go daemon/daemon.go daemon/networkdriver/bridge/driver.go daemon/state.go vendor/src/github.com/docker/libnetwork/ipallocator/allocator.go
Restore failed if network resource not released during checkpoint, e.g., a container with port open with -p Signed-off-by: Hui Kang <[email protected]> Conflicts: daemon/container.go
8f69143
to
59c375a
Compare
5dda3d1
to
183488f
Compare
e91c518
to
988a915
Compare
b584b5a
to
a6a4511
Compare
7c96921
to
7fda470
Compare
d80f2fb
to
9272300
Compare
Docker-DCO-1.1-Signed-off-by: Hui Kang [email protected]