-
Notifications
You must be signed in to change notification settings - Fork 40
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
Add support for 1:1 job per topology assignment #36
Add support for 1:1 job per topology assignment #36
Conversation
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: danielvegamyhre The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
fa97fee
to
537742a
Compare
…y due to CRD size limit for kubectl apply
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
few nits
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
Co-authored-by: Abdullah Gharaibeh <40361897+ahg-g@users.noreply.github.com>
/label tide/merge-method-squash |
/lgtm |
Fixes #27
Adds
Exclusive
struct to the API which has aTopologyKey
field. If specified, the JobSet controller will append pod affinity and anti-affinity terms to each job pod template spec, which ensures 1:1 job per topology assignment, and ensures all pods for each job land on the same topology.Note: I also had to update the
make install
command in the Makefile to usekubectl create
instead ofkubectl apply
as a workaround for this issue.Did a manual test on a cluster with 3 node pools, using the topology key
cloud.google.com/gke-nodepool
to place 1 job per node pool, and it worked.Test JobSet with 3 replicated jobs, 1 replica each:
Pod placement:
Test JobSet with 1 replicated job with 3 replicas:
Pod placement: