Bump containerd client #36684
Bump containerd client #36684
Conversation
libcontainerd/client_daemon.go
Outdated
There was a problem hiding this comment.
We probably need to catch connection failures in normal requests and trigger a reconnect, but we aren't handling these cases at all (or ever AFAIK).
There was a problem hiding this comment.
Those requests would just fail until the connection monitor reconnects normally.
|
Will this change be part of the upcoming containerd 1.1.0 release? Wondering if we can get back to released versions of the dependency (or at least something on a release branch) |
|
I think so, a release branch has not be cut yet. |
|
The new client API is not really "new" anymore since it's been a couple of weeks, btw. |
|
Alright, so probably safe to assume it’ll be in 1.1.0, and once that’s released, we could switch to a tagged version, thanks! |
cpuguy83
force-pushed
the
bump_containerd_client
branch
from
271b782 to
5554c31
Compare
Codecov Report
@@ Coverage Diff @@
## master #36684 +/- ##
=========================================
Coverage ? 35.33%
=========================================
Files ? 614
Lines ? 46345
Branches ? 0
=========================================
Hits ? 16375
Misses ? 27797
Partials ? 2173 |
cpuguy83
force-pushed
the
bump_containerd_client
branch
from
5554c31 to
f7a5424
Compare
This does not bump the containerd binary. Picks last commit before go1.10 switch, which is not currently supported in moby. Signed-off-by: Brian Goff <[email protected]>
This fixes an issue where the containerd client is cached in a container object in libcontainerd and becomes stale after containerd is restarted. Signed-off-by: Brian Goff <[email protected]>
cpuguy83
force-pushed
the
bump_containerd_client
branch
from
f7a5424 to
2c682d5
Compare
|
LGTM |
6 participants
Reconnect()API to reconnect cached clients instead of replacing them (leaving stale clients in other cached objects).