Implement ExecutorUpgradeController
xref https://github.com/mesosphere/kubernetes-mesos/issues/324
Since this controller is supposed to delete pods I had a peek in upstream.
Upstream has methods for deleting pods in the node controller, see https://github.com/kubernetes/kubernetes/blob/master/pkg/controller/node/nodecontroller.go#L727, https://github.com/kubernetes/kubernetes/blob/master/pkg/controller/node/nodecontroller.go#L763.
@jdef @sttts I am tempted to copy these fragments to the upgrade controller, but a generalization could make sense, what do you think?
A 100 lines sound like reasonable large. If it doesn't make the code completely ugly, I would try to factor that out.
After looking at the code, the only dependency to the node controller are the kubeclient and some event sink. Looks sensible to factor out,
@sttts k, I can go that route.
@s-urbaniak please xref the working code you have so far
@jdef it's here, sorry for the delay https://github.com/mesosphere/kubernetes/compare/c8f1019c1064773727f1adccee275dfd3e1d00e8...mesosphere:sur-722-executor-upgrade-controller
thanks!
On Fri, Jan 8, 2016 at 2:16 PM, Sergiusz Urbaniak [email protected] wrote:
@jdef https://github.com/jdef it's here, sorry for the delay mesosphere/kubernetes@ c8f1019...mesosphere:sur-722-executor-upgrade-controller https://github.com/mesosphere/kubernetes/compare/c8f1019c1064773727f1adccee275dfd3e1d00e8...mesosphere:sur-722-executor-upgrade-controller
— Reply to this email directly or view it on GitHub https://github.com/mesosphere/kubernetes-mesos/issues/722#issuecomment-170097185 .
- connected to #324
bah! I thought waffle.io would do something special if I used the "connected to" keyword. disappointed...
@jdef @s-urbaniak @sttts
This is a great feature and will enable running of production workloads on k8sm since upgrades can be done without a complete outage. Very interested in running/testing this out.
Is there any ETA on when this will be complete/merged?
thanks for your interest! currently in bug-fixing mode, it may be a little while before we can circle back around to this
On Fri, Feb 19, 2016 at 8:11 PM, Yash Parikh [email protected] wrote:
@jdef https://github.com/jdef @s-urbaniak https://github.com/s-urbaniak @sttts https://github.com/sttts
This is a great feature and will enable running of production workloads on k8sm since upgrades can be done without a complete outage. Very interested in running/testing this out.
Is there any ETA on when this will be complete/merged?
— Reply to this email directly or view it on GitHub https://github.com/mesosphere/kubernetes-mesos/issues/722#issuecomment-186472972 .