argo VS Orkestra

Like Argo Workflows?

https://github.com/argoproj/argo-workflows

Dagger looks more comparable to Argo Workflows: https://argoproj.github.io/argo-workflows/ That's the first of the Argo projects, which can run multi-step workflows within containers on Kubernetes.

For what it's worth, my colleagues and I have had great luck with Argo Workflows and wrote up a blog post about some of its advantages a few years ago: https://www.interline.io/blog/scaling-openstreetmap-data-wor...

Despite other alternatives like Circle CI, Travis CI, GitLab CI or even self-hosted options using open-source projects like Tekton or Argo Workflow, the reason for choosing GitHub Actions was straightforward: GitHub Actions, in conjunction with the GitHub ecosystem, offers a user-friendly experience and access to a rich software marketplace.

author seems to be describing the kind of patterns you might make with https://argoproj.github.io/argo-workflows/ . or see for example https://github.com/couler-proj/couler , which is an sdk for describing tasks that may be submitted to different workflow engines on the backend.

it's a little confusing to me that the author seems to object to "pipelines" and then equate them with messaging-queues. for me at least, "pipeline" vs "workflow-engine" vs "scheduler" are all basically synonyms in this context. those things may or may not be implemented with a message-queue for persistence, but the persistence layer itself is usually below the level of abstraction that $current_problem is really concerned with. like the author says, eventually you have to track state/timestamps/logs, but you get that from the beginning if you start with a workflow engine.

i agree with author that message-queues should not be a knee-jerk response to most problems because the LoE for edge-cases/observability/monitoring is huge. (maybe reach for a queue only if you may actually overwhelm whatever the "scheduler" can handle.) but don't build the scheduler from scratch either.. use argowf, kubeflow, or a more opinionated framework like airflow, mlflow, databricks, aws lamda or step-functions. all/any of these should have config or api that's robust enough to express rate-limit/retry stuff. almost any of these choices has better observability out-of-the-box than you can easily get from a queue. but most importantly.. they provide idioms for handling failure that data-science folks and junior devs can work with. the right way to structure code is just much more clear and things like structuring messages/events, subclassing workers, repeating/retrying tasks, is just harder to mess up.

Argo Workflows + Argo Events

Drone, or Tekton, Argo Workflows if you’re on k8s

Check out Argo Workflows.

go-git has a lot of bugs and is not actively maintained. The bug even affects Argo Workflow, which caused our data pipeline to fail unexpectedly (reference: https://github.com/argoproj/argo-workflows/issues/10091)

Github Repo - Azure/orkestra: Orkestra is a cloud-native release orchestration and lifecycle management (LCM) platform for the fine-grained orchestration of inter-dependent helm charts and their dependencies (github.com)

Orchestration of a group of complex applications with dependencies between themselves and their subcharts/dependencies is hard

Imagine a scenario in which application that depends on Prometheus, Grafana and Istio. Now let's take it a step further and define a dependency between Istio and it's observability components, Prometheus & Grafana, since Istio itself leverages both Prometheus and Grafana for it's control-plane. In other words we now have a dependency DAG where, the application depends-on Istio depends-on Prometheus & Grafana. Now imagine the application depending on other applications that provide PaaS like infrastructure capabilities required to run the application itself

Defining this dependency order allows all components to spin up both, reliably and efficiently

This would make a huge difference when we are dealing with complex applications like 5G network functions running mission-critical workloads across the network and their lifecycle management (installs/upgrades/rollback). Providers, tend to take releases in a "semi-continuous" manner as opposed to true CD. Meaning, when upgrading, multiple applications and their subcharts may be upgraded in a single release. This is a scenario in which defining the dependency ordering between the applications and their subcharts allows the service providers to limit the blast radius if this go wrong, allowing them to fail fast and rollback without impacting all upgrade-ready apps

Introducing Azure Orkestra https://github.com/Azure/Orkestra that provides the aforementioned capabilities using a familiar Kubernetes declarative CRDscalled ApplicationGroups. Orkestra leverages Argo Workflows to generate DAG workflows to deploy all applications with an application group (and optionally a DAG workflow for the subcharts in each application). The DAG on execution generates HelmRelease objects that are then acted upon by flux helm-operator to reliably perform the helm actions (install, delete, upgrade, rollback)

Introducing Azure Orkestra https://github.com/Azure/Orkestra that provides the aforementioned capabilities and features using familiar Kubernetes declarative objects (CRDs) called ApplicationGroups.