An error occurred while loading the file. Please try again.
-
jorgee authored183f446a
COMP Superscalar Framework ChangeLog
Release number: 1.1.1
Release date: 1-Oct-2013
-------------------------------
This is the first public release of the COMP Superscalar Framework.
Release number: 1.1.2
Release date: 5-Jun-2014
-------------------------------
* C/C++ binding.
* Python binding.
* Integrated Development Environment for COMPSs applications (IDE)
* Priority tasks.
* New tracing system using the Extrae tool.
* Deleting a file within an OE removes all the replicas in the system.
* Updated the SSH Trilead adaptor libraries to remove unnecessary sleeps.
* Scripts for submission to queue systems (LSF, PBS, Slurm).
* Configuration of application directory and library path in project XML file.
* Separate logs for resubmitted / rescheduled tasks.
* Create a COMPSs sandbox in the workers instead of JavaGAT's.
Release number: 1.2
Release date: Nov-2014
-------------------------------
* N implementations for task methods, each with its own constraints.
* Constraint-aware resource management.
* Support for multicore tasks.
* Pluggable schedulers: facilitate the addition of new schedulers and policies.
* Extended support for objects in C/C++ binding.
* Extended IDE for N implementations and deployment through PMES.
* Update cloud connector for rOCCI to work with rocci-cli v4.2.5.
* Enhance rOCCI connector to compute the real VM creation times.
* Extended resources schema to support Virtual Appliances pricing.
* New LSF GAT adaptor.
* Deprecated Azure and EMOTIVE Cloud connectors.
* Deprecated Azure GAT adaptor.
Release number: 1.3
Release date: Nov-2015
-------------------------------
New features:
* Runtime:
- Persistent workers: workers can be deployed on computing nodes and persist during all the application lifetime, reducing runtime overhead.
Previous implementation of workers based on a per task process is still supported.
- Enhanced logging system
- Interoperable communication layer: different inter-nodes communication protocol is supported by implementing the Adaptor interface (JavaGAT
and NIO implementations already included)
- Simplified cloud connectors interface
- JClouds connector
* Python:
- Added constraints support
- Enhanced methods support
- Lists accepted as a tasks' parameter type
- Support for user decorators
* Tools:
- New monitoring tool: with new views, as workload and possibility of visualizing information about previous runs
- Enhanced Tracing mechanism
* Simplified execution scripts
* Simplified installation on Supercomputers
7172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140
Known Limitations:
* Exceptions raised from tasks are not handled by the master
* Java tasks must be declared as public
* Java objects MUST be serializable or, at least, follow the java beans model
* Support limited to SOAP based services
* Persistent Workers do NOT isolate task executions in a sandbox
Release number: 1.4
Release date: April-2016
-------------------------------
New features:
* Runtime:
- Support for Dockers added
- Support for Chameleon added
- Object cache for persistent workers
- Improved error management
- Connector for submitting tasks to MN supercomputer from external COMPSs applications added
- Bug-fixes
* Python:
- Bug-fixes
* Tools:
- Enhanced Tracing mechanism:
· Reduced overhead using native java API
· Support for communications instrumentation added
· Support for PAPI hardware counters added
Known Limitations:
* When executing python applications with constraints in the cloud the initial VMs
must be set to 0
Release number: 2.0 Amapola (Poppy)
Release date: November-2016
-------------------------------
New features:
* Runtime:
- Upgrade to Java 8
- Support to remote input files (input files already at workers)
- Integration with Persistent Objects
- Elasticity with Docker and Mesos
- Multi-processor support (CPUs, GPUs, FPGAs)
- Dynamic constraints with environment variables
- Scheduling taking into account the full tasks graph (not only ready tasks)
- Support for SLURM clusters
- Initial COMPSs/OmpSs integration
- Replicated tasks: Tasks executed in all the workers
- Explicit Barrier
* Python:
- Python user events and HW counters tracing
- Improved PyCOMPSs serialization. Added support for lambda and generator parameters.
* C:
- Constraints support
* Tools:
- Improved current graph visualization on COMPSs Monitor
Improvements:
- Simplified Resource and Project files (NO retrocompatibility)
- Improved binding workers execution (use pipes instead of Java Process Builders)
- Simplifies cluster job scripts and supercomputers configuration
- Several bug fixes
Known Limitations:
* When executing python applications with constraints in the cloud the initial VMs
must be set to 0