Sprint 37
- Will end 12/22
- Sprint 38 will start 1/2
- Kevin is no longer available for the project (someone left SDSC and he's filling in until position filled)
Sprint 37 update
- Quickstart under review
- Cauth annotations for internal SSO under review
- go-client update in progress, with good progress
- Discussed targeting 1.1 release before break
- Quickstart + internal SSO
- Splitting of UI
- Update data page to include NBI quickstart link
- 2.0 release
- Terraform
- Kubernetes 1.8 upgrade + technical debt
- Globus auth
- MDF prototype planned
- Other issues:
- Jetstream outage since 12/7 – ETK instance is offline and no ETA on fix
- SGI webinar on 12/13
- Cybertraining NSF proposal
Sprint 38 planning
- Focus: 2.0 features
- Kubernetes 1.8 + Terraform + Calico + Security policies
- Technical debt (unit/integration testing, refactoring)
- Globus auth
- Catalog redesign
- Gluster security problem +Security policies
- Multiple Workbench instances in one K8s
- etcd > Mongo
- Other data links
- MHD – requires exposing the data, which is complicated
- TERRA-REF – requires upgrading the TERRA-REF workbench, which would be good
- Carma would require exposing the data
- MDF - requires forge environment
Discussion of multi-zone/federation
- Kubernetes federation or application-level federation
- Data synchronization
- Treat backup differently
- Gluster can synchronzie
- Federation vs failover
- Robust external loadbalancer
Discussion of training vs DataDNS:
- TAC was positive on continued work toward DataDNS vision
- However, recent NSF proposals have focused more on training/education capability
- Is there someone we can talk to?
Other considerations:
- Landing page
- Sphinx/Jekyll for consistent documentation. WT + CiS are adopting Sphinx
- Status and outage information
- Do we need integration with JIRA
- Gonzo/Monitoring