|Version 30 (modified by walt, 12 months ago)|
This is a dynamic list of projects recently conducted, currently under way, and/or planned for the near future. Each project page describes the project, technical issues involved, and progress. This is both a place for developers to share ideas and users to both learn about developments to come and provide input. To suggest a post on the developer's list or email one of the developers.
These projects are currently in development. These pages describe the motivation behind the project, and highlight some of the design issues as they are encountered. It is hoped these will also serve as a basis for documenting the development effort and producing future detailed documentation.
- Server to server communication?
- Access control
- Distributed directories
- Data mirroring - on immutable
- Metadata mirroring and migration?
- Searchable metadata
- Distributed server-supported data indexing?
- High performance user interface
- Log users accessing each server?
- Support for remove and list extended attributes?
- Shared client cache - D, N, and/or A?
- aio/lio for direct interface?
These are ideas that are not yet being actively developed but are, for the most part, of particular interest to the OrangeFS community. These may serve as a basis for graduate student research, summer intern projects and future efforts of the OrangeFS development team. We invite discussion on any of these projects.
Features and Performance
- Atomic operations on attributes
- Gather/Scatter Broadcast/Reduce
- State Machine Plug-ins
- Event Plug-ins
- Audit, MD5 Hash on File create, counters
- Background Tasks
- Develop complete set of file utilities?
- File Views?
- tunable ucache
- Restore multi-lookup in security enhanced code?
- Enhanced mmap support?
- Add default layout to configuration and directories?
- java client/interface? for easy integration with Hadoop MR
- Memory management framework (gFramework)?
- Protocol Definition Language
- Remove asserts/add error handling?
- Real-time logging for debugging?
- threading (BMI, metaops, client)?
- Add ability to fork clients - use pthread_atfork()
- Compare readdirplus's way of gathering data size information against tree logic. Modify readdirplus, if tree logic is faster?