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.

Sign up for GitHub

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

Jump to bottom

Make it clear that what a training job is composed of #575

Closed

wangkuiyi opened this issue Jan 31, 2018 · 1 comment

Assignees

wangkuiyi commented Jan 31, 2018

a Kubernetes job of trainers
a Kubernetes replica set of parameter servers
a Kubernetes replica set of the master -- we use the replica set to (1) make sure that there is one and only one master process in a job, and (2) recover it when it fails.

We might need a data type to represent this fact.

wangkuiyi self-assigned this

Collaborator

gongweibao commented Feb 1, 2018

we use the replica set to (1) make sure that there is one and only one master process in a job

we use etcd to
(1) make sure that there is one and only one master process in a job and (2) recover it when it fails.

wangkuiyi mentioned this issue

Refactorize EDL controller and autoscaler #579

Merged

helinwang closed this as completed in #579

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment