quantum-serverless/gateway
Pablo Aragón 3efde15a19
feat: job provider list endpoint (#1583)
* feat: job provider list endpoint

* fix lint

* apply partial review suggestions

* fix provider-endpoint

* change serializer for provider_list

* fix lint
2025-02-20 13:09:29 -05:00
..
api feat: job provider list endpoint (#1583) 2025-02-20 13:09:29 -05:00
main update-version-0.19.0 (#1580) 2025-02-04 12:20:41 -05:00
probes Gateway: fair share scheduling and resource limitation (#570) 2023-06-05 09:50:21 -04:00
scripts Gateway: fair share scheduling and resource limitation (#570) 2023-06-05 09:50:21 -04:00
templates Remove old deprecated actions (#1391) 2024-06-26 15:04:14 -04:00
tests feat: job provider list endpoint (#1583) 2025-02-20 13:09:29 -05:00
version add version api endpoint to gateway (#1279) 2024-04-09 16:09:15 -04:00
.dockerignore Updated dockerignore from the root (#429) 2023-04-12 19:04:20 +02:00
.gitignore APIs improvements and docker-compose fixes (#399) 2023-04-12 18:26:09 +02:00
.pylintrc Issue 254 | Gateway service (#274) 2023-03-21 14:39:39 -04:00
Dockerfile Include in gateway scarf configuration (#1582) 2025-02-10 09:09:55 -05:00
README.md Hot fix for client test (#1478) 2024-08-30 11:26:36 -04:00
entrypoint.sh Sunset keycloak from the project (#1134) 2023-12-07 10:15:27 -05:00
manage.py Rename Quantum Serverless to Qiskit Serverless (#1309) 2024-05-07 15:11:03 -04:00
requirements-dev.txt Update dependency pylint-django to >=2.5.5 (#1098) 2023-11-16 13:46:00 -05:00
requirements.txt Update cryptography version (#1585) 2025-02-18 11:42:40 -05:00
tox.ini simplify tox ini (#1558) 2024-12-17 13:12:00 -05:00
utils.py Rename Quantum Serverless to Qiskit Serverless (#1309) 2024-05-07 15:11:03 -04:00

README.md

Qiskit Serverless gateway - test

Gateway is a set of apis that are used as a backend for providers.

Build image

docker build -t qiskit/qiskit-serverless/gateway:<VERSION> .

Env variables for container

Variable Description
DEBUG run application on debug mode
SITE_HOST host of site that will be created for Django application
RAY_HOST Host of Ray head node that will be assigned to default created compute resource
DJANGO_SUPERUSER_USERNAME username for admin user that is created on launch of container
DJANGO_SUPERUSER_PASSWORD password for admin user that is created on launch of container
DJANGO_SUPERUSER_EMAIL email for admin user that is created on launch of container
SETTINGS_TOKEN_AUTH_URL URL for custom token authentication
SETTINGS_TOKEN_AUTH_USER_FIELD user field name for custom token authentication mechanism. Default userId.
SETTINGS_TOKEN_AUTH_TOKEN_FIELD user field name for custom token authentication mechanism. Default apiToken.
SETTINGS_AUTH_MECHANISM authentication backend mechanism. Default mock_token. Options: mock_token and custom_token. If custom_token is selected then SETTINGS_TOKEN_AUTH_URL must be set.
SETTINGS_TOKEN_AUTH_VERIFICATION_URL URL for custom token verificaiton
SETTINGS_TOKEN_AUTH_VERIFICATION_FIELD name of a field to use for token verification
RAY_KUBERAY_NAMESPACE namespace of kuberay resources. Should match kubernetes namespace
RAY_NODE_IMAGE Default node image that will be launched on ray cluster creation
RAY_CLUSTER_MODE_LOCAL 0 or 1. 1 for local mode (docker compose), 0 for cluster mode where clusters will be created by kuberay
RAY_CLUSTER_MODE_LOCAL_HOST if RAY_CLUSTER_MODE_LOCAL set to 1, then this host for ray head node will be used to run all workloads
LIMITS_JOBS_PER_USER number of concurrent programs/jobs user can run at single point of time
LIMITS_MAX_CLUSTERS number of compute resources can be allocated in single point of time
RAY_CLUSTER_TEMPLATE_CPU default compute kuberay template cpu setting
RAY_CLUSTER_TEMPLATE_MEM default compute kuberay template memory setting
RAY_CLUSTER_WORKER_REPLICAS worker replicas per cluster
RAY_CLUSTER_WORKER_REPLICAS_MAX maximum number of worker replicas per cluster
RAY_CLUSTER_WORKER_MIN_REPLICAS min worker replicas per cluster for auto scaling
RAY_CLUSTER_WORKER_MIN_REPLICAS_MAX maximum number of min worker replicas per cluster for auto scaling
RAY_CLUSTER_WORKER_MAX_REPLICAS max replicas per cluster for auto scaling
RAY_CLUSTER_WORKER_MAX_REPLICAS_MAX maximum number of max worker replicas per cluster for auto scaling
RAY_CLUSTER_MAX_READINESS_TIME max time in seconds to wait for cluster readiness. Will fail job if cluster is not ready in time.
QISKIT_IBM_CHANNEL Channel that will be set in env variables in jobs for QiskitRuntimeService client
QISKIT_IBM_URL Authentication url for QiskitRuntimeService that will be set for each job