Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
- Home
- Archived Content
- Azure Service Fabric Blog
- Release of SDK 2.3.301 and Runtime 5.3.301 for Windows
Release of SDK 2.3.301 and Runtime 5.3.301 for Windows
- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Printer Friendly Page
- Report Inappropriate Content
Published
Aug 06 2019 03:58 PM
395
Views
Aug 06 2019
03:58 PM
Aug 06 2019
03:58 PM
First published on MSDN on Oct 13, 2016
Today, we are releasing Service Fabric SDK v2.3.301 and Runtime v5.3.301 for Windows. This release includes a number of new features, along with a set of important bug fixes. Because of the critical nature of the runtime fixes you should upgrade your clusters to the latest runtime version immediately. If you are using the standalone version of Service Fabric or are have moved to manual upgrades of Service Fabric in Azure, make sure you complete the upgrade at the earliest. For those customers who are hosted on Azure and have chosen automatic upgrades, we have initiated the upgrade to the latest version already.
The SDK install fails on clean installations of Windows 10 Anniversary edition. For more details refer to the "SDK fails to install on Windows 10 Anniversary edition" at the bottom of the post.
All cluster versions prior to 5.3.121 will be out of support on January 20 th , 2017. Due to an issue related to Node Id generation as described in the release notes , all newly created Service Fabric clusters, with the runtime version 5.2.207 and 5.3.121 are at the risk of going in to an unrecoverable state. Hence, we strongly recommend that you upgrade your cluster to the latest version of the Service Fabric runtime (version 5.3.301) at the earliest. This guidance also applies to version 5.3.204 of the standalone runtime release, which should also be upgraded as soon as possible.
The Service Fabric tools for Visual Studio are not yet compatible with recent changes to the installer for Visual Studio “15” preview. As a result, we are not providing those tools with this release. They will be added in an upcoming release.
You can now choose from a set of supported versions of the Service Fabric runtime on your Azure cluster and control when runtime upgrades are applied. By default, clusters will continue to be upgraded to the latest release automatically so action is required if you wish to switch to the manual upgrade mode.
For more details, see Upgrade a Service Fabric Cluster in Azure .
Note: This feature introduces the possibility that the runtime in your Service Fabric cluster may be lower than the latest supported runtime version. Please see the “Supported Runtime Versions” section above.
As with the Azure Service, you can choose the runtime version you want to upgrade to for Service Fabric on Windows Server. For more information, see Upgrading a standalone Windows Server Cluster .
You can now specify the application capacity parameters individually (previously you were required to pass all parameters). Starting with this release you can specify the individual parameters in the form of {tag:value}.
Previously, fault injection and chaos testing were performed using the client-side testability APIs surfaced through the ChaosTestService type in System.Fabric.Testability.Scenario namespace. Thus a failure in the client resulted in a failure in the test. Now, you can invoke fault injection and chaos tests using the built-in FaultAnalysisService.
We recently announced a set of new build and release primitives in Visual Studio Team Services designed to make continuous integration and continuous delivery of Service Fabric applications easier. With this release, it is simpler to configure that environment directly from the Visual Studio client.
A Service Fabric node is identified and addressed in the system by a NodeId. Service Fabric runtime generates the NodeId from the name of the node. There was a bug in this algorithm which can cause a node to come up with a different NodeId upon restart with one of the two possible values; one of which is incorrect. This could result in a non-existent node showing up in PowerShell queries or SFX with the same name but a different NodeId. If a replica was hosted on the node while it had an incorrect id, that replica will be lost. Service Fabric is designed to tolerate loss of nodes and replicas with minimal availability loss. However, this adds a new failure mode into the system and increases the probability of the cluster becoming unhealthy on subsequent failures. This bug affects newly created clusters in Service Fabric version 5.2.207 and 5.3.121. All released versions of standalone GA are affected.
Creating a large number of ServiceProxy or ActorProxy instances can cause high memory usage (e.g., when creating a new instance every time a remote procedure call is made). This release includes a fix that prevents high memory build-up when creating a large number of these objects.
We made several small changes to the CRM.Decision traces to improve the information provided and reduce the volume of traces.
Due to an issue with the Service Fabric lease layer driver, the SDK fails to install on clean installations on Windows 10 Anniversary edition. For development machines only, use the following links to install the 5.2.207 version as a stop-gap, if disabling secure boot isn't possible. The installation must proceed in order - Runtime first, SDK second and Tools third.
5.2.207 Runtime: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftServi...
5.2.207 SDK: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftServi...
5.2.207 Tools: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftAzure...
For information about other new features and bug fixes included in the 5.3.301 release, please see the detailed release notes .
Cheers,
The Service Fabric Team
Today, we are releasing Service Fabric SDK v2.3.301 and Runtime v5.3.301 for Windows. This release includes a number of new features, along with a set of important bug fixes. Because of the critical nature of the runtime fixes you should upgrade your clusters to the latest runtime version immediately. If you are using the standalone version of Service Fabric or are have moved to manual upgrades of Service Fabric in Azure, make sure you complete the upgrade at the earliest. For those customers who are hosted on Azure and have chosen automatic upgrades, we have initiated the upgrade to the latest version already.
- Runtime, SDK, and tools for Visual Studio 2015 (requires Visual Studio 2015 Update 2 or later)...
- Runtime and SDK only (no Visual Studio tools)
- Service Fabric for Windows Server: Upgrade your standalone cluster for Windows Server
The SDK install fails on clean installations of Windows 10 Anniversary edition. For more details refer to the "SDK fails to install on Windows 10 Anniversary edition" at the bottom of the post.
Supported Runtime Versions
All cluster versions prior to 5.3.121 will be out of support on January 20 th , 2017. Due to an issue related to Node Id generation as described in the release notes , all newly created Service Fabric clusters, with the runtime version 5.2.207 and 5.3.121 are at the risk of going in to an unrecoverable state. Hence, we strongly recommend that you upgrade your cluster to the latest version of the Service Fabric runtime (version 5.3.301) at the earliest. This guidance also applies to version 5.3.204 of the standalone runtime release, which should also be upgraded as soon as possible.
The Service Fabric tools for Visual Studio are not yet compatible with recent changes to the installer for Visual Studio “15” preview. As a result, we are not providing those tools with this release. They will be added in an upcoming release.
Highlights
Control the cluster version
You can now choose from a set of supported versions of the Service Fabric runtime on your Azure cluster and control when runtime upgrades are applied. By default, clusters will continue to be upgraded to the latest release automatically so action is required if you wish to switch to the manual upgrade mode.
For more details, see Upgrade a Service Fabric Cluster in Azure .
Note: This feature introduces the possibility that the runtime in your Service Fabric cluster may be lower than the latest supported runtime version. Please see the “Supported Runtime Versions” section above.
As with the Azure Service, you can choose the runtime version you want to upgrade to for Service Fabric on Windows Server. For more information, see Upgrading a standalone Windows Server Cluster .
Improved API for specifying application capacity parameters in PowerShell
You can now specify the application capacity parameters individually (previously you were required to pass all parameters). Starting with this release you can specify the individual parameters in the form of {tag:value}.
Fault injection and the cluster analysis service are now built in
Previously, fault injection and chaos testing were performed using the client-side testability APIs surfaced through the ChaosTestService type in System.Fabric.Testability.Scenario namespace. Thus a failure in the client resulted in a failure in the test. Now, you can invoke fault injection and chaos tests using the built-in FaultAnalysisService.
Configure continuous integration and continuous delivery using Visual Studio Team Services from Visual Studio
We recently announced a set of new build and release primitives in Visual Studio Team Services designed to make continuous integration and continuous delivery of Service Fabric applications easier. With this release, it is simpler to configure that environment directly from the Visual Studio client.
Important Bug Fixes
A same Node can appear with two different Node IDs
A Service Fabric node is identified and addressed in the system by a NodeId. Service Fabric runtime generates the NodeId from the name of the node. There was a bug in this algorithm which can cause a node to come up with a different NodeId upon restart with one of the two possible values; one of which is incorrect. This could result in a non-existent node showing up in PowerShell queries or SFX with the same name but a different NodeId. If a replica was hosted on the node while it had an incorrect id, that replica will be lost. Service Fabric is designed to tolerate loss of nodes and replicas with minimal availability loss. However, this adds a new failure mode into the system and increases the probability of the cluster becoming unhealthy on subsequent failures. This bug affects newly created clusters in Service Fabric version 5.2.207 and 5.3.121. All released versions of standalone GA are affected.
Memory leak in Service and Actor Remoting
Creating a large number of ServiceProxy or ActorProxy instances can cause high memory usage (e.g., when creating a new instance every time a remote procedure call is made). This release includes a fix that prevents high memory build-up when creating a large number of these objects.
Reducing Cluster Resource Manager Trace Volume
We made several small changes to the CRM.Decision traces to improve the information provided and reduce the volume of traces.
SDK fails to install on Windows 10 Anniversary edition
Due to an issue with the Service Fabric lease layer driver, the SDK fails to install on clean installations on Windows 10 Anniversary edition. For development machines only, use the following links to install the 5.2.207 version as a stop-gap, if disabling secure boot isn't possible. The installation must proceed in order - Runtime first, SDK second and Tools third.
5.2.207 Runtime: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftServi...
5.2.207 SDK: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftServi...
5.2.207 Tools: http://download.microsoft.com/download/9/E/3/9E31A03B-755C-4C14-89FC-1844E1C80E37/MicrosoftAzure...
For information about other new features and bug fixes included in the 5.3.301 release, please see the detailed release notes .
Cheers,
The Service Fabric Team
0
Likes