There is a ManageIQ Sprint review every 3 weeks, and everyone in the community is welcome to participate. To join our Sprint review meeting, please use this Bluejeans link.
For Sprint 43, we merged 329 pull requests, making this a regular sprint averaging 100 PRs per week. There is a good distribution between new features, bug fixes, and code refactoring. We have plenty of enhancements on the Providers side, including Amazon, Networking, and Hawkular. Work has started on a Service Broker, which is a new piece for ManageIQ to allow VM selection based on criteria you provide.
SmartState has been extended with a generalized disk LRU caching module, which will be employed to address Azure performance and throttling issues. PostgreSQL has been upgraded to 9.5. There are also many User Interface updates, explained in detail in the following video recording of the meeting, including a couple of demos in the Providers area (Recheck Authentication and Credentials Validation).
As we had audio issues during this meeting with people dropping in and out, some of the presentations were not in the order laid out in the agenda. Here is the actual order with the corresponding timestamps for each section:
- Sprint Statistics (Oleg Barenboim) - 0:53
- Community Update (Carol Chen) - 4:54
- Providers (Bronagh Sorota, Alissa Bonas, Greg McCullough) - 8:38
- Automate (Greg McCullough, Madhu Kanoor) - 14:35
- User Interface (Dan Clarizio) - 21:27
- Platform (Gregg Tanzillo) - 35:12
- Service Broker (Chris Kacerguis) - 38:35
- SmartState (Rich Oliveri) - 39:14
- Performance (Dennis Metzger) - 41:17
- REST API (Alberto Bellotti) - separate video
On June 30, we released Darga 2 containing mainly bug fixes and stabilization. With already over 200 PRs backported, we'll continute to stabilize Darga 3 aimed for an end-August release.
In case you missed it, we also started a new blog series, "Last week in ManageIQ", where select pull requests from the codebase are featured weekly. As it's impractical to go through all the 300+ contributions during Sprint reviews, this is a great way to check out some highlights each week. The first post was published this Tuesday. We hope you'll find it interesting and useful!
Sprint 44 review will be on August 3, 2016 @ 7:30am PDT/10:30am EDT/14:30 GMT. You can import the ManageIQ community calendar to be notified about this and future Sprint reviews.
- Service resolution based on Provision Order
- Added /System/Process/MiqEvent instance
- Added Provider refresh call to Amazon retire state machine in Pre-Retirement state
- Service Dialogs: Added ‘Visible’ flag to all dialog fields
- Class Schema allows for EMS, Host, Policy, Provision, Request, Server, Storage, and VM (or Template) datatypes
- Value is id of the objects
- If object is not found, the attribute is not defined.
- Null Coalescing Operator
- Multiple String values separated by “||”
- Evaluated on new attribute data type “Null Coalescing”
- Order dependent, left to right evaluation
- First non-blank value is used
- Skip and warn about missing objects
- Platform: Custom attributes support in reporting and expressions
- Selectable as a column
- Usable in report filters and charts
- Amazon: Public Images Filter
- Separate Google Network Manager
- NFV: VNFD Templates and VNF Stacks
- Deployment entity operations: Undeploy, redeploy
- Server operations: Reload, suspend, resume
- Live metrics for datasources and transactions
- Performance reports for middleware servers
- Support for alert profiles and alert automated expressions for middleware server
- Crosslink middleware servers with RHEV VMs
- Collect and display deployment status
- Datasources topology view
- REST API
- Support for report schedules
- Support for approving or denying service requests
- Support for OpenShift Container Deployments
- Support for Virtual Templates
- Service Broker
- Started work on Service Broker to allow ManageIQ to select VM for you based on criteria (cloud, cost, or performance)
- Added API backend for Resourceless Servers
- Added datastore for the default settings for resourceless servers
- SmartState: Generalized disk LRU caching module
- Caching module can be used by any disk module, eliminating duplication.
- Can be inserted “higher” in the IO path.
- Configurable caching parameters (memory vs performance)
- Will be employed to address Azure performance and throttling issues.
- Other disk modules converted over time.
- User Interface
- Settings moved to top right navigation header
- C3 Charts fully implemented - chart interaction coming soon!
- Tagging for Ansible Tower job templates
- Live Search added to bootstrap selects
- Self Service UI Order History with detail
- Automate: Generic Object: Model refactoring/cleanup, use PostgreSQL jsonb column
- Capacity and Utlization improvements included reduced number of SQL queries and number of objects
- Improved tag processing for Alert Profiles
- UI Performance: specific pages targeted resulted in up to 98% reduction in rendering for those pages
- Platform: PostgreSQL upgraded to 9.5 needed for HA feature
- Pluggability: Began extraction of Amazon Provider into separate repository
- Upgraded to hawkular gem version 2.2.1
- Refactor infrastructure for easier configuration
- Ansible: Automate method updated to pass JobTemplate “Extra Variables” defined in the Provision Task
- User Interface
- Default Filters tree converted to TreeBuilder - more on the way
- Cloud Key Pair form converted to AngularJS (Dana - UX team)
- Toolbars:Cleaned up partials, YAML -> classes
- Provider Forms: Credentials Validation improvements
This is a companion discussion topic for the original entry at http://manageiq.org/blog/2016/07/sprint-43-review-july-13-2016/