What is the expected enhancement?

Update documentation for classes, function and tutorials

What is the expected enhancement?

As we are doing with redis and ray let's see if we can remove our jupyter dependency in favor of artifactory.

Setup github actions to run tests for client and manager packages.

Notes: as this project in private at this moment we have constraints on actions time. Probably we need to think about how to reduce actions time, while maintaining code quality

What is the expected behavior?

Integrate https://github.com/keycloak/keycloak as access management tool to provide access to services within deployment.

Job server should have JWT auth

Issues in project:

• #136
• #137
• #138

What is the expected enhancement?

Add documentation on deployment

What is the expected enhancement?

In the infrastructure project we have several references where we are using middleware where we should use manager.

What is the expected enhancement?

Add optional redis chart to our helm setup

What is the expected behavior?

Use chart repository to store helm charts

What is the expected enhancement?

Jobs: add support of long running jobs

Ray have an SDK for jobs https://docs.ray.io/en/latest/cluster/running-applications/job-submission/sdk.html

For initial iteration we will use this feature

Example:

serverless = QuantumServerless()

job_id = serverless.run_job(entrypoint="python  job.py")

serverless.get_job_logs(job_id)

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Repository problems

These problems occurred while renovating this repository.

• WARN: Job timeout
• WARN: Worker exiting

Rate-Limited

These updates are currently rate-limited. Click on a checkbox below to force their creation now.

• Update dependency requests to >=2.28.1
• Update Terraform aws to >= 3.75.2
• Update Terraform aws to >= 4.40.0
• Update Terraform github.com/aws-ia/terraform-aws-eks-blueprints to v4.17.0
• Update Terraform hashicorp/terraform to >= 1.3.5
• Update Terraform ibm to >= 1.47.1
• Update Terraform kubernetes to >= 2.16.0
• Update dependency Flask to v2.2.2
• Update dependency importlib-metadata to >=4.13.0
• Update dependency qiskit-ibm-runtime to >=0.8.0
• Update dependency qiskit-nature to >=0.5.1
• Update dependency qiskit-terra to >=0.22.2
• Update dependency ray to >=2.1.0, <3
• Update dependency werkzeug to v2.2.2
• Update lachlanevenson/k8s-kubectl Docker tag to v1.25.0
• Update python Docker tag to v3.11
• Update rayproject/ray Docker tag to v2.2.0
• Update actions/checkout action to v3
• Update actions/setup-python action to v3
• Update actions/setup-python action to v4
• Update actions/upload-artifact action to v3
• Update dependency flask-restx to v1
• Update dependency importlib-metadata to v5
• 🔐 Create all rate-limited PRs at once 🔐

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Update Terraform helm to >= 2.7.1
• Update dependency qiskit-aer to >=0.11.1

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

What is the expected enhancement?

Update docker images for nodes, manager, jupyter

What is the expected enhancement?

Allow to quantum_serverless library implements a events-telemetry class to reuse between different programs that can run in different pods.

We can start with something simple (key, value), like:

quantum_serverless.events.notify(key, value) # Will be broadcast to all listeners listen by key
@quantum_serverless.events.listener(key) # Internally key will be the combination of env+key

What is the expected enhancement?

Split docker compose into 2 version:

1. dev version which builds images from file and current project
2. version with images from qiskit dockerhub with latest commit for user to play around

Something like this

Dev

services:
  jupyter:
    container_name: qs-jupyter
    build:
      context: ./
      dockerfile: infrastructure/docker/Dockerfile-notebook

Regular

services:
  jupyter:
    container_name: qs-jupyter
    image: qiskit/quantum-serverless-notebook:latest

What is the expected enhancement?

Add get_job_by_id method to QuantumServerless object

Details

Currently when you execute run_job it will return a job object that you can query. We need to get job object by id to recover it in a case of you lost job variable

What is the expected enhancement?

New version of runtime has been released yesterday and has some breaking changes.
Update to latest version and make sure everything works as expected.

What is the expected enhancement?

There are at least 2 types of ports we need to support:

1. interactive (client) mode for 10001
2. dashboard/jobs for 8265

We now only supporting one port for client mode

In compute resource setting we need to support both ports, but make them with default values.

What is the expected behavior?

For the project we need to provide to the user an initial configuration with what be able to start running things in different clouds.

With this issue we will provide the configuration for IBM Cloud.

What is the expected behavior?

Create a new interface called NestedPrograms to allow encapsulate the run_qiskit_remote methods in a class to bring more powerful logic regarding to reuse/import/export the code in the near future.

The proposal can be something like:

# It will have all/by type default libraries imported

class NestedProgram (...): # To extend in your class
	… # local parameters like visibility : ENUM …

	def __init__(…, env, options): # Options allowed to define the resources to use
		…

	def _import_program(…): # Code imported to run. From local/remote. Overwritten
		…

	def program(in: quantum_serverless.INPUT): # Code to run. Overwritten
		…
		return (quantum_serverless.OUTPUT) out

	@quantum_serverless.run_qiskit_remote(…)
	def run(…, options)�		…
		# Run the nested program in the secure and controlled environment
		output = self.program(…)
		…

Then, we can create workflows with the new interface, allowing something like

# Invoking Nested Programs

env = { ‘name’:’my_nested_environment_customized’, ’options’: { … } } # It can be a VAR ENV?

myProgram = new MyNestedProgram(env, …)
result_step_1 = myProgram.run(…)


myProgram2A = new NestedProgram(env, {‘from’: ‘cloud’, ‘origin’: ‘ENDPOINT’, ‘options’: { … } }))
future_result_2A = myProgram2A.runAsync(result_step_1)

myProgram2B = new NestedProgram(env, {‘from’: ‘cloud’, ‘origin’: ‘ENDPOINT’, ‘options’: { … } }))
future_result_2B = myProgram2B.runAsync(result_step_1)

[result_2A, result_2B] = quantum_serverless.get([future_result_2A, future_result_2B])

results_step_2  = combine_results_step2(result_2A, result_2B)


myProgram3 = new NestedProgram(env, {‘from’: ‘file’, ‘origin’ ‘PATH|URL’, ‘options’: {…}))
final_result = myProgram3.run(results_step_2)

What is the expected enhancement?

Docs: add state to groundstate demo tutorial

What is the expected enhancement?

Allow to quantum_serverless library implements a memory state class to reuse between different programs that can run in different pods.

We can start with something simple (key, value), like:

quantum_serverless.state.set(key, value) # Atomic operation
@quantum_serverless.state.get(key, value) # Internally k will be the combination of env+k, think about it

Migrate helm chart to https://github.com/ray-project/kuberay as it has more mature set of features rather standard one in ray project itself.

What is the expected enhancement?

Update markdown (in repository) documentation in /infrastructure directory.

What is the expected enhancement?

Update markdown (in repository) documentation in /manager directory.

What is the expected behavior?

Add stress test program to test scalability on given compute resources.

Create a program that combines lot of small and big tasks into a big computational graph to check compute resource scaling. Sort of a benchmark program.

What is the expected enhancement?

Add example to readme with usage of runtime primitives

What is the expected behavior?

Now that we are going to start delivering infrastructure resources to users we should start closing a bit the versions related with the dependencies that we are using. In this case Jupyter notebook.

What is the expected enhancement?

Update markdown (in repository) documentation in /client directory.

What is the expected enhancement?

Kuberay helm chart provides ApiService to manage ray clusters (CRUD).

We need to migrate to this helm chart and update client to work with those resources.

Issues in project:

• #130
• #131
• #132
• #133

What is the expected enhancement?

Compute resources is a more general term and should be used instead of clusters.

QServerless object should have access to multiple providers and each provider should be able to have multiple compute resources.

Additional notes

Remove context, set_clusters, cluster, backends

Steps to reproduce the problem

Run any computation within serverless context.

What is the current behavior?

Following deprecation warning appears

/home/ray/anaconda3/lib/python3.7/site-packages/ray/_private/worker.py:983: UserWarning: len(ctx) is deprecated. Use len(ctx.address_info) instead.',
 '  warnings.warn("len(ctx) is deprecated. Use len(ctx.address_info) instead.")

What is the expected behavior?

Should not raise deprecation warning.

How to fix suggestions?

Remove usage of Contex class in type hinting for QuantumServerless class.

What is the expected enhancement?

Add nightly build and push for docker images using GH actions

What is the expected enhancement?

In the terraform project we need to specify in the folder of each provider the values for the helm execution. Here we should review:

• Unify naming:
  • IBM Cloud: terraform.yaml
  • AWS: values.yaml
• Review values: it seems we are using old configurations to run the helm.

What is the expected behavior?

Setup IBM cloud public registry where all images will be pushed to

What is the current behavior?

With the current helm configuration the manager chart is creating two kind of roles: ClusterRole and the default Role.

What is the expected behavior?

ClusterRole is no longer needed and can be removed from the configuration.

What is the expected enhancement?

Add support for get for only finished jobs

Example

from quantum_serverless import get_finished

...

with serverless:
      refs = [run_something() for _ in range(50)]
      time.sleep(2)

      finished_after_2_secods = get_finished(refs)
      print(len(finished_after_2_secods))
# 19

What is the expected enhancement?

Docs: add full getting started documentation with local setup (docker-compose)

What is the expected enhancement?

Implement new interfaces

Init

import quantum_serverless as qs
from quantum_serverless import QuantumServeless

serverless = QuantumServerless()
# or
serverless = QuantumServerless(config)
# or
serverless = QuantumServerless().load_config(...)

Annotation

@qs.run_qiskit_remote(target={"cpus": 2})
def func(...):
    ...

serverless = QuantumServerless():
with serverless.provider("ibm"):
    remote_function_future = func(...)

Steps to reproduce the problem

https://github.com/Qiskit-Extensions/quantum-serverless/actions/runs/3926302643/jobs/6711892304

What is the current behavior?

Nightly build is failing due to manager build process which is involving helm chart.

What is the expected behavior?

Manager should not depend on helm chart or should pull form registry during build process

Steps to reproduce the problem

Helm install with changing cluster values

What is the current behavior?

Not taking values into account

Solution

Fix typo in default values file.

High level of cluster configs should be ray-cluster, not cluster

What is the expected enhancement?

Quantum-Serverless has several dependencies that we are going to monitor in case new versions are published to maintain the project correctly updated.

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Rate-Limited

These updates are currently rate-limited. Click on a checkbox below to force their creation now.

• Update dependency Django to >=4.2.13
• Update dependency nbsphinx to >=0.9.4
• Update dependency tqdm to >=4.66.4
• Update alpine/k8s Docker tag to v1.30.0
• Update dependency dj-rest-auth to >=5.1.0
• Update dependency django-allauth to >=0.62.1
• Update dependency djangorestframework to >=3.15.1
• Update dependency nbqa to >=1.8.5
• Update dependency requests-mock to >=1.12.1
• Update dependency responses to >=0.25.0
• Update dependency s3fs to >=2023.12.2
• Update dependency setuptools to >=69.5.1
• Update dependency sphinx-autodoc-typehints to >=1.25.3
• Update dependency tox to >=4.15.0
• Update grafana/loki Docker tag to v2.9.8
• Update grafana/promtail Docker tag to v2.9.8
• Update icr.io/quantum-public/quantum-serverless-ray-node Docker tag to v0.10.1
• Update opentelemetry packages to >=1.24.0 (opentelemetry-api, opentelemetry-exporter-otlp, opentelemetry-exporter-otlp-proto-grpc, opentelemetry-sdk)
• Update prom/prometheus Docker tag to v2.51.2
• Update python Docker tag to v3.11.4
• Update qiskit packages (qiskit-ibm-provider, qiskit-ibm-runtime)
• Update registry.access.redhat.com/ubi8/openssl Docker tag to v8.9-13
• Update Helm release nginx-ingress-controller to v10
• Update Helm release nginx-ingress-controller to v11
• Update Helm release postgresql to v14
• Update Helm release postgresql to v15
• Update actions/checkout action to v4
• Update actions/setup-python action to v5
• Update actions/upload-artifact action to v4
• Update azure/setup-helm action to v2
• Update azure/setup-helm action to v3
• Update azure/setup-helm action to v4
• Update dependency Django to v5
• Update dependency Sphinx to v7
• Update dependency black to v23
• Update dependency black to v24
• Update dependency cloudpickle to v3
• Update dependency coverage to v6
• Update dependency coverage to v7
• Update dependency cryptography to v42.0.7
• Update dependency dj-rest-auth to v6
• Update dependency importlib-metadata to v6
• Update dependency importlib-metadata to v7
• Update dependency kubernetes to v27
• Update dependency kubernetes to v28
• Update dependency kubernetes to v29
• Update dependency mypy to v1
• Update dependency mypy-extensions to v1
• Update dependency pylint to v3
• Update dependency pytest to v7
• Update dependency pytest to v8
• Update dependency pytest-randomly to v2
• Update dependency pytest-randomly to v3
• Update dependency reno to v4
• Update dependency s3fs to v2024
• Update dependency sphinx-autodoc-typehints to v2
• Update dependency testcontainers to v4
• Update docker/build-push-action action to v4
• Update docker/build-push-action action to v5
• Update docker/login-action action to v3
• Update docker/setup-buildx-action action to v3
• Update docker/setup-qemu-action action to v3
• Update grafana/loki Docker tag to v3
• Update grafana/promtail Docker tag to v3
• Update peaceiris/actions-gh-pages action to v4
• Update peter-evans/create-pull-request action to v6
• Update release-drafter/release-drafter action to v6
• 🔐 Create all rate-limited PRs at once 🔐

Other Branches

These updates are pending. To force PRs open, click the checkbox below.

• Update dependency ipython to >=8.24.0

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Update dependency cryptography to >=41.0.7
• Update Helm release kuberay-operator to v1.1.1
• Update dependency ray to v2.20.0
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

What is the expected enhancement?

Add specific packages installation guide.

This is happening through parameters of run_qiskit_remote decorator.

What is the expected behavior?

Setup GH actions workflow on release.

On release:

• compile images
• push images to IBM registry
• build client
• push to PyPi

What is the expected enhancement?

Update groundstate solver demo / tutorial to latest nature and to run locally on any estimator

What is the expected behavior?

Job: add ability to cancel/stop job

Should be something like this

job = serverless.run_program(program)

job.stop() # or job.cancel()

What is the expected behavior?

As QuantumServerless is fully compatible with Ray we can reuse ray datasets to illustrate an example of usage distributed datasets to work with quantum problems

What is the expected enhancement?

Update docker compose with notebook and head node for quick demonstrations.

What is the expected enhancement?

Kuberay team just started to upload their charts to https://github.com/ray-project/kuberay-helm

Let's remove ray charts from repo and use this artifactory

What is the expected enhancement?

Add integration tests for jobs.

This will require to setup of running head node inside docker container and testing functionality against actual running head node instead of local one.

What is the expected enhancement?

Update parallel_sweep_solver vqe with Primitives.

Reference material to update VQE is in this notebook https://github.com/mberna/qce22-qiskit-runtime-tutorial/blob/main/docs/tutorials/Building%20algorithms%20with%20Qiskit%20Runtime.ipynb