Skip to content

Sprint Summaries 2017

Jump to bottom
Nicolas Degory edited this page Jan 11, 2018 · 42 revisions

Sprint 25 - December 13, 2017

  • Released v0.18.0 CANCELED
    • Protected Docker socket (TLS)
    • Core images based on Alpine 3.7 (security)
    • Documentation update
    • Bug fixes

Ongoing/next sprint

  • Security epic
  • ARS support
  • reach out ASS team for observability integration

Sprint 24 - November 30, 2017

  • Released v0.17.0
  • Draft REST API documentation
  • Pagination in logs API
  • Object storage API / CLI command
  • Cluster status and cluster node list implemented in the API
  • Extended cluster status in the CLI output
  • gRPC gateway updated to 1.3.0
  • CLI accepts AWS credential as arguments (alternative method)
  • CLI error messages more user-friendly
  • CLI service inspect subcommand now accepts the --format option
  • Fixed the following issues:
    • Alerts rules sample for Prometheus 2.0
    • Prometheus link in alert manager
    • CLI deletion of a rollbacked AWS stack
    • Force cleanup of docker volumes at cluster destruction
    • Automatic creation of the index pattern for Kibana
    • Race condition between system prune and cluster creation

Ongoing/next sprint

  • Security:
    • Based on first TwistLock scan of the AMP images, upgrade to Alpine 3.7
    • Docker TLS
    • SSH access restrictions on AWS
  • Cluster deployment User Experience and Documentation

Sprint 23 - November 16, 2017

  • Released v0.16.0
  • Prototype deployment of APIRS on AMP is almost complete: deployment configuration needs further review by APIRS team (https://github.com/appcelerator/amp/tree/master/examples/stacks/apirs)
  • Updated to Docker 17.09 stable release
  • Improved security with network isolation
  • Improved stack deployment (registry authentication, environment variables)
  • Improved CLI to reflect more accurate statuses for tasks, services and stacks
  • Refactored AMP metrics API to use Prometheus API instead of Elasticsearch's
  • Added option to provision NFS on AWS cluster deployment
  • Updated various versions:
    • Golang 1.9
    • Prometheus 2.0.0
    • Grafana 4.6.1
    • nats 0.6.0
  • Significantly reduced technical debt

Ongoing/next sprint

  • Security:
    • Run Twistlock checks on AMP images
    • Run Fortify & ThreadFix
  • Provide cloud deployment storage solution (i.e. S3 on AWS)

Sprint 22 - November 2, 2017

  • Switched to Prometheus for user services metrics acquisition. #1730, #1729. A new default Grafana dashboard has been added accordingly:
  • Cluster wide NFSv4 endpoint #1739
  • Started a prototype deployment of APIRS on AMP using stack files confluence/APIRS+on+AMP
  • Fixes and enhancements:
    • Stack deployment API now support environment variables #1737
    • user signup robustness #1736
    • cloud cluster creation with CLI now displays the progress #1732

Ongoing/next sprint

  • Prototype deployment of APIRS on AMP JIRA/AMP-4
  • AMP metrics API refactoring to use Prometheus API instead of Elasticsearch's JIRA/AMP-21
  • Node management with InfraKit JIRA/AMP-24

Sprint 20 & 21 - October 19, 2017

Ongoing/next sprint

  • Prototyping a deployment of APIRS using stacks on top of AMP.

Sprint 19 - September 25, 2017

  • Most of the architecture work has been completed and published here.
  • Study about Integration of AMP/ArrowCloud with AxwayID is still in progress. Required help from the architecture team.
  • Nikolai is giving a hand on the migration of ArrowVPC this week.
  • Nicolas is on vacation for the next 2 weeks.
  • Joseph is now working part-time and start working on Azure deployments.

Ongoing/next sprint

  • Working with the architecture team on Axway ID integration.

Sprint 18 - September 11, 2017

  • Starting to work with Wei to define the best strategy to be able to deliver ArrowCloud on top of AMP
    • Compare architectures and define a plan/roadmap to support this target architecture
  • Tony has left on sabbatical
  • Nikolai to work remotely (from Ukraine)
  • François is no longer part of AMP
  • Joseph will join part-time by end of September

Ongoing/next sprint

  • Work with Wei and her team
  • Nicolas will be on vacation

Sprint 17 - Aug 28, 2017

Ongoing/next sprint

  • Maintenance / fixes
  • Housekeeping: docs / refactoring / code clean up / vendor updates
  • Team outreach
  • Roadmap planning and technical R&D

Sprint 16 - Aug 14, 2017

Ongoing/next sprint

  • Maintenance / fixes
  • Housekeeping: docs / refactoring / code clean up / vendor updates
  • Team outreach
  • Roadmap planning and technical R&D

Sprint 15 - July 31, 2017

Next sprint

  • Architectural / technical planning
  • Housekeeping: docs / refactoring / code clean up / vendor updates
  • Team outreach

Sprint 14 - July 17, 2017

  • Numerous fixes, usability, and stability updates
    • Security model refactoring
    • Fixed issues related to resource access management
    • Bash and Zsh completion for the CLI
  • Updates to docs website: http://appcelerator.io/
  • Significant progress on AWS cluster deployment
    • Releasing 0.13 on Wednesday (July 17)
    • Stable version of AWS deployment

Next sprint

  • wrap AWS cluster deployment
  • enhance docs, create videos
  • team outreach

New AMP landing page / docs site:

amp-landing-page

Sprint 13 - July 3, 2017

  • Releasing 0.12 on Wednesday (July 5)
    • Stable version for local deployment

Next sprint

  • focus on AWS cluster deployment

Sprint 12 - June 19, 2017

Releasing this sprint (targeting Monday).

Video Links

Sprint 11 - June 5, 2017

  • Many fixes

General

  • Added service ls, ps and inspect commands
  • Added super user and super organization
  • Fixed logs ordering
  • Added a stack status
  • Added a more detailed service status (reporting if placement constraints don't match any node)
  • Added custom dashboards

Cluster

  • HA and self healing Cluster deployment on AWS
  • Cluster health alerting
  • UI for user and team management, stack and service management
  • Stack file filtering for reserved resources

UI

  • User organization status management (Member or Owner)
  • Custom dashboard
  • Dashboard as a resource sharable with authorization
  • Super user management / super org

Sprint 10 - May 22, 2017

Released 0.10.1 last week

  • Add CLI for Windows (amp.exe)
  • Faster local deployment (uses host swarm mode instead of a DinD cluster)
  • Email verification can be disabled for local develpment and testing
  • Updated docs and examples
  • Fix file permission issues for builds on Linux
  • Simplifies the developer local workflow (flags for registration and notifications are not required by default)
  • Synchronizes all the images to the same tag

Release 0.11.0 today

  • Significant new functionality in web UI
  • Various stability enhancements

https://github.com/appcelerator/amp/releases

Notes

arrow-reference-server-on-amp

getting-started-with-arrow-on-amp

Sprint 9 - May 8, 2017

  • Released 0.9.1 patch last Wednesday, 0.10.0 tomorrow
    • Most improvements are "under the hood"
    • Significant stability/usability/performance enhancements ranging over
      • local cluster deployments
      • improved developer experience on linux
      • vendor package updates
      • examples and documentation updates
      • HTTP REST gateway for gRPC API; all HTTP -> HTTPS
      • Web UI enhancements (Francois to demo)
    • API / command UX 95%+ complete
      • Focus next sprint is on isolated swarms
      • Only two more sprints until our baseline release

Sprint 8 - April 24, 2017

  • Released 0.9.0 in time for DockerCon
    • Significant milestone release
    • Hosted version at cloud.atomiq.io
    • Team spent last week at DockerCon
    • Goals for 1.0 include improvements around securing and sharing resources, provisioning additional cluster/swarms isolated from primary, reverse proxy enhancements, UI/UX, documentation and examples
    • DEMO
      • amp
      • pinger
      • visualizer
      • web ui

Sprint 7 - April 10, 2017

Preparing MVP release for DockerCon

  • Continuing to update/refactor for final preparing mvp by DockerCon
  • Hosted version at cloud.atomiq.io
  • Bertrand will demo using CLI to connect
  • Team will continue testing/fixing/refactoring this week and will be at DockerCon next week (except Francois).

Sprint 6 - March 27, 2017

  • French team members have (finally) arrived...
...last week and are getting settled.

  • AMP Teams with RBAC support has been merged
Big effort last week merging in all the amp teams/security code from the branch that Bertrand, Francois, Neha, and Joseph had been working on.

  • Local and remote clusters are now symmetric
InfraKit implementation previously demonstrated for AWS is now working for local development, replaces my custom implementation, and is integrated into the CLI for bootstrapping a local cluster and docker swarm environment.

  • Preparing MVP release for DockerCon
The codebase is going through a thorough scrubbing and still being updated/refactored as we make a final big push for the version we are going to release during the week that ends with the new sprint in time for DockerCon; this will be a fully functional version of the platform that includes cluster provisioning, hosting, stack deployments/updates/rollbacks, monitoring, logging, and teams support with role-based access controls.

Sprint 4 - February 27, 2017

  • Our Docker InfraKit based provisioning strategy works for AWS. Nicolas will demo.
  • AMP authentication and basic authorization is in place, as well as email and SMS support for account verification, reset password, etc. Bertrand has prepared a demo.
http://bit.ly/ampauth
  • The migration / refactoring for Docker 1.13 swarm mode and compose v3 files is still ongoing and is expected to be complete next sprint (Sprint 5).

Next sprint

  • Nicolas will work with Joseph to extend our InfraKit-based cluster provisioning strategy across multiple cloud providers.
  • Bertrand will work with Neha on the permissions model for granular role-based access control over resources.
  • I will work with Francois on a number of other areas related to refactoring around Docker 1.13 . I will highlight this functionality at the next sprint demo.

Notes

  • I expect to have the full team here in Silicon Valley for the following sprint (Sprint 6) as we work to prepare AMP for 1.0 and production.
  • The version of AMP that we release next month (before DockerCon) will be a significant milestone and we will then begin the process of working with PSG to meet the organizations Strategic Security Bar.

Sprint 3 - February 13, 2017

  • Docker 1.13 and compose v3 file support refactoring
  • More improvements to toolchain - go executables to wrap dockerized process
  • Bertrand taking the lead on the Account model (org, team, resources permission support)
  • Working on cluster deployments using Docker InfraKit
  • Email support for account verification, password resets, etc
  • CLI standardization improvements
  • Test engine enhancements

Next sprint

  • Continued work on the 1st 4 items above

Notes

  • Personnel changes, offer being made

Sprint 2 - January 30, 2017

  • Docker 1.13 swarm and compose v3 file support refactoring
  • Message Queue package
  • Data layer and CLI support for resource authorization
  • Containerized toolchain
    • resolved a long-running permissions issue
    • streamlined makefile
    • improved tool images)

Next sprint

  • Continue work on all of the above

Notes

  • Olivier Pignault, Director of Product Portfolio Management (reports to Vince Padua, Head of Innovation) will start working with us going forward
  • Francois' wife's going into surgery and will be out this week
Clone this wiki locally