Skip to main content

quapp-hpc

Python library cho Quapp HPC functions — cầu nối giữa Quapp FaaS platform và Slurm HPC cluster.

Architecture

ksvc (Docker)
├── index.py                  FastAPI server
├── quapp_hpc/
│   ├── factory/
│   │   └── hpc_handler_factory.py   Entry point cho user
│   ├── component/backend/
│   │   └── hpc_invocation.py        Orchestrates job lifecycle
│   └── model/
│       ├── provider/slurm_provider.py   Auth headers, base URL
│       └── device/slurm_device.py       Submit → Poll → S3 download
└── function/
    └── handler.py            User-defined processing() + post_processing()

Luồng thực thi

index.py nhận HTTP POST
    → HpcHandlerFactory.create_handler(event, processing_fn, post_processing_fn)
    → InvocationHandler.handle()
    → HpcInvocation.submit_job()
        1. processing_fn(invocation_input)  → bash script string
        2. SlurmDevice._create_job(script)  → POST Slurm REST API → slurm_job_id
        3. SlurmDevice._get_job_result()    → poll mỗi 30s → COMPLETED/FAILED
        4. SlurmDevice._download_s3_result()→ boto3 get s3://$S3_BUCKET/$JOB_UUID/output.json
        5. post_processing_fn(s3_result)    → final response

Environment variables (từ K8s Secret slurm-credentials)

Var Ví dụ Mô tả
SLURM_API_URL http://10.1.0.15:6820 Slurm REST API base URL
SLURM_JWT eyJ... JWT token cho Slurm auth
SLURM_USERNAME quapp-svc Slurm username
SLURM_ACCOUNT quapp Slurm account/allocation
S3_BUCKET quapp-slurm-output-dev S3 bucket cho job output
AWS_REGION ap-southeast-1 AWS region
SLURM_POLL_SEC 30 Polling interval (giây)
SLURM_TIMEOUT_SEC 21600 Max wait time (giây, default 6h)

invocation_input schema

Xem chi tiết tại ../qapp-sdk-templates/slurm-hpc/README.md.

Tóm tắt:

{
  "resources": { "partition", "nodes", "cpus_per_task", "gpus", "memory_gb", "time_limit" },
  "container": { "type": "sif"|"docker"|"none", "image": "..." },
  "job":       { "type": "script"|"command", "script"|"command": "...", "environment": {}, "input_s3_paths": [] }
}

Slurm REST API

  • Version: v0.0.40
  • Submit: POST {SLURM_API_URL}/slurm/v0.0.40/job/submit
  • Status: GET {SLURM_API_URL}/slurm/v0.0.40/job/{job_id}
  • Auth headers: X-SLURM-USER-NAME, X-SLURM-USER-TOKEN

Job state mapping

Slurm state Quapp state
PENDING, CONFIGURING, RUNNING, COMPLETING RUNNING
COMPLETED DONE
FAILED, CANCELLED, TIMEOUT, NODE_FAIL, PREEMPTED ERROR

S3 output pattern

Job script phải upload kết quả:

aws s3 cp /tmp/output.json s3://$S3_BUCKET/$JOB_UUID/output.json

$JOB_UUID$S3_BUCKET được inject tự động bởi SlurmDevice._create_job() qua Slurm environment array.

K8s Secret

# infrastructure/quapp-job-scheduler/k8s/cts/slurm-secret.yaml
apiVersion: v1
kind: Secret
metadata:
  name: slurm-credentials
  namespace: quapp-functions-dev
stringData:
  SLURM_JWT: "<generate: sudo scontrol token username=quapp-svc lifespan=2592000>"
  SLURM_API_URL: "http://10.1.0.15:6820"
  SLURM_USERNAME: "quapp-svc"
  SLURM_ACCOUNT: "quapp"
  S3_BUCKET: "quapp-slurm-output-dev"
  AWS_REGION: "ap-southeast-1"

DB seed required

Chạy script trước khi deploy:

infrastructure/quapp-functions-backend/docs/db/seed_slurm_hpc.sql

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

quapp_hpc-0.0.1.tar.gz (15.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

quapp_hpc-0.0.1-py3-none-any.whl (18.4 kB view details)

Uploaded Python 3

File details

Details for the file quapp_hpc-0.0.1.tar.gz.

File metadata

  • Download URL: quapp_hpc-0.0.1.tar.gz
  • Upload date:
  • Size: 15.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.10.12

File hashes

Hashes for quapp_hpc-0.0.1.tar.gz
Algorithm Hash digest
SHA256 05e9e545e524d1e6b063eafba428da214a41c2bc6fc9af5f6caf879a195f9051
MD5 8519a6141cd6625a3bc69be22b9d7dc8
BLAKE2b-256 4e033f4a8d09d0f98ae726f31f6dacba1d3b69bdc67c608bca684aa703be3fcb

See more details on using hashes here.

File details

Details for the file quapp_hpc-0.0.1-py3-none-any.whl.

File metadata

  • Download URL: quapp_hpc-0.0.1-py3-none-any.whl
  • Upload date:
  • Size: 18.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.10.12

File hashes

Hashes for quapp_hpc-0.0.1-py3-none-any.whl
Algorithm Hash digest
SHA256 019bc12516aff865b7a6dcf384009b7afe52b3ac019d31994f864026627694ee
MD5 9f11677d14c405a647a0ade1e7f305f5
BLAKE2b-256 ae75f6896c45b1e4d406ecf00eb13c2c5a67cccc2fb2c5c09fd06a37c1217695

See more details on using hashes here.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page