Releases: thecodeteam/scaleio-framework
Releases · thecodeteam/scaleio-framework
ScaleIO Framework v0.3.1
Version 0.3.1 (2017/02/28)
ScaleIO Framework 0.3.1 introduces a couple of new experimental features.
Changes since 0.3.0:
ScaleIO Framework v0.3.0
Version 0.3.0 (2017/01/10)
ScaleIO Framework 0.3.0 introduces the new Imperative approach for deploying ScaleIO.
Changes since 0.2.0:
- Addressed Issue #82: Implement User Defined Deployment aka Imperative Deployment. Please see PR #111 for more details.
- Addressed Issue #95: Create a RHEL7 Version of Cloud Formation Template that can be used for demos, testing, etc. RHEL7 will be the preferred platform for development and testing.
- Another feature that has been enhanced is that when new nodes are brought online, storage devices will be added to the ScaleIO cluster based on the method being used.
- If the Imperative approach is used, only devices defined by the user will be added.
- If the existing automatic method is used, any available block devices that are currently not being used (ie has a filesystem on it) will be added to the default domain/pool.
- Addressed Issue #71: Instead of using elastic ips, each mesos node has been configured to dynamically modify required files (such as hostname, etc) on boot and then start the mesos services.
- Addressed Issue #103: Full Framework documentation is available on Read the Docs.
ScaleIO Framework v0.3.0-rc2
Version 0.3.0-rc2 (2016/12/19)
ScaleIO Framework 0.3.0 introduces the new Declarative approach for deploying ScaleIO.
Changes since 0.2.0:
- Addressed Issue #82: Implement User Defined Deployment. Please see PR #111 for more details.
- Addressed Issue #95: Create a RHEL7 Version of Cloud Formation Template that can be used for demos, testing, etc. RHEL7 will be the preferred platform for development and testing.
- Another feature that has been enhanced is that when new nodes are brought online, storage devices will be add to the ScaleIO cluster based on the method being used.
- If the Declarative approach is used, only devices defined by the user will be added.
- If the existing automatic method is used, any available block devices that are currently not being used (ie has a filesystem on it) will be added to the default domain/pool.
- Addressed Issue #71: Instead of using elastic ips, each mesos node has been configured to dynamically modify required files (such as hostname, etc) on boot and then start the mesos services.
ScaleIO Framework v0.3.0-rc1
Version 0.3.0-rc1 (2016/12/15)
ScaleIO Framework 0.3.0 introduces the new Declarative approach for deploying ScaleIO.
Changes since 0.2.0:
- Addressed Issue #82: Implement User Defined Deployment. Please see PR #111 for more details.
- Another feature that has been enhanced is that when new nodes are brought online, storage devices will be add to the ScaleIO cluster based on the method being used.
- If the Declarative approach is used, only devices defined by the user will be added.
- If the existing automatic method is used, any available block devices that are currently not being used (ie has a filesystem on it) will be added to the default domain/pool.
ScaleIO Framework v0.2.0
Version 0.2.0 (2016/11/10)
Changes since 0.1.0:
- Addressed Issue #65: RHEL7 and CentOS7 Support. Supports ScaleIO 2.0.1
- Addressed Issue #91: Updated Ubuntu14 to support ScaleIO 2.0.1. The CloudFormation template in the demo folder has also been updated to handle ScaleIO 2.0.1.
- Fixed Issue #93: The REX-ray configuration file that is created follows the suggested best practices.
- Massive restructuring to the executor. This was largely in part due to time to market release of 0.1.0. With addition of RHEL7/CentOS7 support, the project needed to be restructured to support multiple platforms in a maintainable fashion.
- Supports Mesos Master leader changes. Implements the Mesos Master redirect functionality to connect to a different master.
- Removed the following flags. This is largely in part to due differences with both DEB and RPM package managers (in command and operational behavior) between versions of platforms (ie RHEL6 vs RHEL7).
- scaleio.deb.mdm
- scaleio.deb.sds
- scaleio.deb.sdc
- scaleio.deb.lia
- scaleio.deb.gw
- scaleio.rpm.mdm
- scaleio.rpm.sds
- scaleio.rpm.sdc
- scaleio.rpm.lia
- scaleio.rpm.gw
- Added platform specific flags for the ScaleIO packages. This is largely in part due to each platform having a different DEB or RPM between platform versions. Added the following flags:
- scaleio.ubuntu14.mdm
- scaleio.ubuntu14.sds
- scaleio.ubuntu14.sdc
- scaleio.ubuntu14.lia
- scaleio.ubuntu14.gw
- scaleio.rhel7.mdm
- scaleio.rhel7.sds
- scaleio.rhel7.sdc
- scaleio.rhel7.lia
- scaleio.rhel7.gw
- Renamed the following 3 flags to match the CPU flags.
- executor.memory.mdm -> executor.mem.mdm
- executor.memory.no -> executor.mem.non
- executor.memoryfactor -> executor.memfactor
- Added an intelligent reboot feature which will fix a reboot timing issue when the ScaleIO node that is running the scheduler is rebooted before other nodes have had the opportunity to contact it for the current state. I have not seen this happen yet, but there was certainly the possibility. That has been resolved now.
- Added a new flag "Debug" to help with debugging the scheduler and executor. Among some of the things the debug flag does is prevent the reboot of the Mesos Agent node.
- Fixed an issue that sometimes caused REX-Ray not to start on reboot. Placed additional dependencies on ScaleIO scini driver.
- Fixed Issues #94 and #72: Documentation related changes.
ScaleIO Framework v0.2.0-rc2
Version 0.2.0-rc2 (2016/11/09)
Changes since 0.1.0:
- Addressed Issue #65: RHEL7 and CentOS7 Support. Supports ScaleIO 2.0.1
- Addressed Issue #91: Updated Ubuntu14 to support ScaleIO 2.0.1. The CloudFormation template in the demo folder has also been updated to handle ScaleIO 2.0.1.
- Fixed Issue #93: The REX-ray configuration file that is created follows the suggested best practices.
- Massive restructuring to the executor. This was largely in part due to time to market release of 0.1.0. With addition of RHEL7/CentOS7 support, the project needed to be restructured to support multiple platforms in a maintainable fashion.
- Supports Mesos Master leader changes. Implements the Mesos Master redirect functionality to connect to a different master.
- Removed the following flags. This is largely in part to due differences with both DEB and RPM package managers (in command and operational behavior) between versions of platforms (ie RHEL6 vs RHEL7).
- scaleio.deb.mdm
- scaleio.deb.sds
- scaleio.deb.sdc
- scaleio.deb.lia
- scaleio.deb.gw
- scaleio.rpm.mdm
- scaleio.rpm.sds
- scaleio.rpm.sdc
- scaleio.rpm.lia
- scaleio.rpm.gw
- Added platform specific flags for the ScaleIO packages. This is largely in part due to each platform having a different DEB or RPM between platform versions. Added the following flags:
- scaleio.ubuntu14.mdm
- scaleio.ubuntu14.sds
- scaleio.ubuntu14.sdc
- scaleio.ubuntu14.lia
- scaleio.ubuntu14.gw
- scaleio.rhel7.mdm
- scaleio.rhel7.sds
- scaleio.rhel7.sdc
- scaleio.rhel7.lia
- scaleio.rhel7.gw
- Renamed the following 3 flags to match the CPU flags.
- executor.memory.mdm -> executor.mem.mdm
- executor.memory.no -> executor.mem.non
- executor.memoryfactor -> executor.memfactor
- Added an intelligent reboot feature which will fix a reboot timing issue when the ScaleIO node that is running the scheduler is rebooted before other nodes have had the opportunity to contact it for the current state. I have not seen this happen yet, but there was certainly the possibility. That has been resolved now.
- Added a new flag "Debug" to help with debugging the scheduler and executor. Among some of the things the debug flag does is prevent the reboot of the Mesos Agent node.
- Fixed an issue that sometimes caused REX-Ray not to start on reboot. Placed additional dependencies on ScaleIO scini driver.
- Fixed Issues #94 and #72: Documentation related changes.
ScaleIO Framework v0.2.0-rc1
Version 0.2.0-rc1 (2016/11/07)
Changes since 0.1.0:
- Fixed Issue #65: RHEL7 and CentOS7 Support. Supports ScaleIO 2.0.1
- Fixed Issue #91: Updated Ubuntu14 to support ScaleIO 2.0.1. The CloudFormation template in the demo folder has also been updated to handle ScaleIO 2.0.1.
- Fixed Issue #93: The REX-ray configuration file that is created follows the suggested best practices.
- Massive restructuring to the executor. This was largely in part due to time to market release of 0.1.0. With addition of RHEL7/CentOS7 support, the project needed to be restructured to support multiple platforms in a maintainable fashion.
- Supports Mesos Master leader changes. Implements the Mesos Master redirect functionality to connect to a different master.
- Removed the following flags. This is largely in part to due differences with both DEB and RPM package managers (in command and operational behavior) between versions of platforms (ie RHEL6 vs RHEL7).
- scaleio.deb.mdm
- scaleio.deb.sds
- scaleio.deb.sdc
- scaleio.deb.lia
- scaleio.deb.gw
- scaleio.rpm.mdm
- scaleio.rpm.sds
- scaleio.rpm.sdc
- scaleio.rpm.lia
- scaleio.rpm.gw
- Added platform specific flags for the ScaleIO packages. This is largely in part due to each platform having a different DEB or RPM between platform versions. Added the following flags:
- scaleio.ubuntu14.mdm
- scaleio.ubuntu14.sds
- scaleio.ubuntu14.sdc
- scaleio.ubuntu14.lia
- scaleio.ubuntu14.gw
- scaleio.rhel7.mdm
- scaleio.rhel7.sds
- scaleio.rhel7.sdc
- scaleio.rhel7.lia
- scaleio.rhel7.gw
- Renamed the following 3 flags to match the CPU flags.
- executor.memory.mdm -> executor.mem.mdm
- executor.memory.no -> executor.mem.non
- executor.memoryfactor -> executor.memfactor
- Added an intelligent reboot feature which will fix a reboot timing issue when the ScaleIO node that is running the scheduler is rebooted before other nodes have had the opportunity to contact it for the current state. I have not seen this happen yet, but there was certainly the possibility. That has been resolved now.
- Added a new flag "Debug" to help with debugging the scheduler and executor. Among some of the things the debug flag does is prevent the reboot of the Mesos Agent node.
- Fixed Issues #94 and #72: Documentation related changes.
ScaleIO Framework v0.1.0
Version 0.1.0 (2016/09/28)
Formal release of 0.1.0
Changes since RC4:
- Updates to the main and demo READMEs with: 1) an example application to deploy, and 2) video demo of the framework in action
- Reboot after install of ScaleIO components decrease from 1 minute to an immediate reboot
- Added a reboot state so its obvious that newly add Agent nodes are in the middle of the reboot process
ScaleIO Framework v0.1.0-rc4
Version 0.1.0-rc4 (2016/09/27)
Changes since RC3:
- Added a reboot to all nodes once ScaleIO, REX-Ray and mesos-module-dvdi have been installed and configured. This gets around an issue in which restarting services due to configuration updates places those processes until the containerization of the executor.
- Updated the REX-Ray init.d to include a dependency on starting the ScaleIO scini driver thus allowing REX-Ray to start normally on reboot
- Added 2 more wait states such that any reboot is coordinated between the existing ScaleIO nodes
- Updated the Amazon CloudFormation template to 1) enable external volumes which is required in order to use external volumes with the Docker containerizer, and 2) moved the work_dir for the master and agents to /usr/lib/mesos in order to debug across reboots
ScaleIO Framework v0.1.0-rc3
Version 0.1.0-rc3 (2016/09/22)
Changes since RC2:
- Added support to the CloudFormation template to support Docker workflows
- Updated the main README.md to be more polished
- Added a separate README for launching the demo. Renamed "aws-demo" folder to "demo".
- Updated the Marathon JSON for the demo to include the default ScaleIO clusterid