Argo Workflows v2.5 Released
v2.5 is the biggest release of Argo Workflows since v2.0 — so let us tell you all about it!
v2.5 in Numbers
More than…
- 29 new features enhancements added
- 74 bug fixed
- 225 pull requests merged
- 60 contributors
12 official new users:
- Ant Financial
- BioBox Analytics
- Capital One
- CCRi
- CyberAgent
- EBSCO Information Services
- Fairwinds
- Greenhouse
- HOVER
- InsideBoard
- Maersk
- Peak AI
- Red Hat
- Sidecar Technologies
Argo UI is dead — long live the Argo Server
In the previous versions, you could use the Argo UI, written in NodeJS, to view your workflows. We’ve completely re-written the Argo UI as a React-based single-page web app with a Golang backend. This is a strategic decision that will allow us to reuse more code within the user interface.
The new UI is not read-only — it also comes with the ability to create and update data directly in your browser.
Workflow Archive
One popular feature request that we’ve for a while now is the need to keep completed workflows for a long time. With v2.5 you can now enable the workflow archive and your completed workflows will be automatically saved into the archive (a MySQL or Postgres database) where you can search them using standard SQL tools, access the results, or re-submit them.
Cron Workflows
One of our most requested features has finally arrived! You can now create Workflows that run on a schedule using our new CronWorkflow CRD. Creating one is simple and it supports all of the Kubernetes `CronJob` options that you’re already familiar with, such as `concurrencyPolicy`, `successfulJobHistoryLimit`, etc. It even has timezone support! CronWorkflows also supports all of Argo’s features, such as WorkflowTemplates.
Take a look: https://github.com/argoproj/argo/blob/master/docs/cron-workflows.md
Brand New APIs
Before v2.5, if you wanted API access to workflows, you could use the Kubernetes API to do this. Unfortunately, this cannot support node status offloading or the workflow archive. We’ve now introduced an API that allows you to interact with the Argo Server over HTTP. This enables the building of new tooling, such as API client, SDKs, DSLs — that use Argo Workflows.
Thanks To Argo Community!
Special thanks go to the Argo community and Argo workflows contributors without whom nothing would have been possible. Our contributor list keeps growing and they are contributing a lot of cool features and enhancement. Many thanks to Argo users for providing feedback on different use cases, testing the RC builds and creating bugs.
