volcano

volcano copied to clipboard

[WIP]Create a new Agent scheduler for Agent workload fast scheduling

2

#### What type of PR is this? feature #### What this PR does / why we need it: Add a new scheduler to execute fast path scheduling for workload like...

qi-min

Gang aware eviction design doc

6

#### What type of PR is this? Propose a design for gang-aware preempt and reclaim. This design is for feature https://github.com/volcano-sh/volcano/issues/4607 #### What this PR does / why we need...

vzhou-p

### What is the problem you're trying to solve Merge the changes from PR #4786 [fix] when update replicaSet, need to synchronize and update the podgroup-related information into the release-1.13...

JesseStutler

Reclaim UTs fail intermittently

2

### Description After [#4774 ](https://github.com/volcano-sh/volcano/pull/4774) was merged, unit tests (UTs) failed sporadically ### Steps to reproduce the issue make unit-test ### Describe the results you received and expected UTs pass...

wangyang0616

Enables role-based elasticity and Gang Scheduling for distributed inference workloads (Kthena)

2

### What is the problem you're trying to solve Each server group of Kthena corresponds to one PodGroup for inference instance-level Gang Scheduling, and each role instance maps to a...

wangyang0616

Support network topology constrain with highestTierName

3

#### What type of PR is this? /kind feature #### What this PR does / why we need it: When submitting a job and specifying the affinity tier, users can...

zhaoqi612

:Draft feat: Add dequeue strategies support for queue scheduling

4

#### What type of PR is this? /kind feature #### What this PR does / why we need it: This PR introduces dequeue strategies for Volcano queues, allowing administrators to...

kingeasternsun

# [Feature] Add Liveness and Readiness Probes to Core Deployments

4

This PR introduces default and customisable `livenessProbe` and `readinessProbe` configurations for the `volcano-admission`, `volcano-controller`, and `volcano-scheduler` deployments. ## Motivation Volcano is a critical control plane component in environments that use...

AxcelBlaze003

add sharding controllers to support node partitioning among multiple scheduler

3

#### What type of PR is this? feature: this PR implements sharding controllers for dynamic node partitioning between schedulers #### What this PR does / why we need it: This...

ssfffss

Optimize CI with reusable image workflow

15

#### What this PR does / why we need it: This PR optimizes the CI pipeline by extracting Docker image building into a centralized reusable workflow. Previously, each E2E test...

Deepam02