Skip to content

Added heterogenous job support#178

Merged
xinli-sw merged 2 commits into
NVIDIA:mainfrom
nvjullin:het-job
May 29, 2026
Merged

Added heterogenous job support#178
xinli-sw merged 2 commits into
NVIDIA:mainfrom
nvjullin:het-job

Conversation

@nvjullin
Copy link
Copy Markdown
Contributor

Heterogenous job allows multiple segments, used for more flexible scheduling and to allow multi-rack deployments.
For simplicity, this PR only splits jobs into one segment each for prefill and decode, instead of one per worker or more elaborate setups.
Use by setting het_jobs: true in resources, default false.

@codecov-commenter
Copy link
Copy Markdown

codecov-commenter commented May 28, 2026

Codecov Report

❌ Patch coverage is 88.65248% with 16 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (main@20f6cde). Learn more about missing BASE report.

Files with missing lines Patch % Lines
src/srtctl/cli/mixins/telemetry_stage.py 63.15% 7 Missing ⚠️
src/srtctl/core/runtime.py 89.18% 4 Missing ⚠️
src/srtctl/core/config.py 66.66% 2 Missing ⚠️
src/srtctl/core/slurm.py 89.47% 2 Missing ⚠️
src/srtctl/cli/do_sweep.py 66.66% 1 Missing ⚠️
Additional details and impacted files
@@           Coverage Diff           @@
##             main     #178   +/-   ##
=======================================
  Coverage        ?   65.72%           
=======================================
  Files           ?       67           
  Lines           ?     8386           
  Branches        ?        0           
=======================================
  Hits            ?     5512           
  Misses          ?     2874           
  Partials        ?        0           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@xinli-sw xinli-sw merged commit babf250 into NVIDIA:main May 29, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants