[dependabot]

Updated Microsoft.Orleans.Sdk from 10.3.1 to 10.4.2.

Release notes

Sourced from Microsoft.Orleans.Sdk's releases.

10.4.2

What's Changed

• [10.x] fix(azure): tolerate legacy membership writers during rolling upgrades by @​ReubenBond in [10.x] fix(azure): tolerate legacy membership writers during rolling upgrades dotnet/orleans#11624

Full Changelog: dotnet/orleans@v10.4.1...v10.4.2

10.4.1

Orleans 10.4.1

Orleans 10.4.1 improves cluster membership, grain scheduling, graceful shutdown, client startup, serialization, and provider reliability. It also restores standalone serializer code generation, fixes ADO.NET streaming message loss, and improves dashboard statistics and navigation.

Compatibility and upgrade notes

[!IMPORTANT]
Public API compatibility: SiloAddress.GetUniformHashCodes(int) now returns ImmutableArray<uint> instead of uint[]. This is a source- and binary-breaking change for code which calls this method. Rebuild affected applications and libraries against 10.4.1. Use ToArray() when a mutable array is required, or Sort() for an immutable sorted result. See #​11425.

• Distributed grain directory: Preserve the cluster's existing partition count and ordered partition boundaries throughout an upgrade. The new GrainDirectoryOptions.GetPartitionBoundaries callback supports configuring the mapping used by an existing cluster, including the older sorted-hash mapping. Partition defaults remain unchanged. See #​11425.
• Management deactivation: Begin using the new IManagementGrain.DeactivateGrain operation after all active silos have been upgraded to support it. See #​11483.
• Custom consistent-ring configurations: Clusters using UseVirtualBucketsConsistentRing = false receive corrected request routing. During a rolling upgrade, older callers can still route reminder operations to the wrong owner. Persisted reminders are reconciled on the owner's next successful refresh, which defaults to five minutes; early occurrences can be missed before reconciliation. Custom grain services using this ring need application-specific upgrade handling. The default virtual-buckets provider retains its existing behavior. See #​11598.

What's changed

The release preparation PR, #​11566, bundles multiple backports. Those changes are expanded below, with links to their original PRs alongside the subsequent 10.x fixes.

Grain execution and graceful shutdown

• Preserve read-only scheduling isolation. Request admission accounts for every active request's interleaving policy, keeping writable requests isolated from active [ReadOnly] requests and correctly handling mixed MayInterleave workloads. #​10809
• Make grain timer scheduling consistent. Timer ticks coalesce while a callback is queued or running, delayed wakeups are preserved, and Change calls made during that interval apply the latest schedule after the callback completes. #​11464
• Carry queued requests with migrating activations. Graceful silo retirement and explicit activation migration transfer eligible queued requests through normal destination admission. Best-effort caller retargeting reduces SiloUnavailableException failures when a retiring silo departs, while preserving callers' original deadlines. Applications should continue to account for ambiguous outcomes when an acceptance acknowledgement is lost. #​11460
• Run observer-level incoming call filters. Filters implemented by observer objects participate in incoming call processing. #​11507
• Safely deactivate existing grain instances. IManagementGrain.DeactivateGrain(GrainId, CancellationToken) locates existing activations directly and waits for their deactivation callbacks. This supports lifecycle operations on grains whose persisted state cannot currently be read. #​11483

Cluster membership, directories, and Kubernetes

• Preserve the default grain directory when registering a named distributed directory. Grains continue to use their configured directory, fixing duplicate activations caused by the registration interaction. #​11553
• Keep membership snapshots canonical and retain departed-silo knowledge. Local startup lifecycle tracking remains separate from published membership rows. Retained death information lets client routing and durable-job adoption account for silos after their rows are pruned. Snapshot decoding also supports older silos which omit the new history field. #​11593
• Improve membership health monitoring. Health monitors handle shutdown cancellation cleanly, and joining silos reevaluate stale peers. #​11421, #​11431
• Upgrade legacy Azure Table membership data safely. Existing legacy tables can be read with their versions and ETags intact. The first successful membership write atomically establishes the snapshot boundary markers, and reads retry when they overlap marker installation. #​11596
• Read Firestore membership rows atomically. The silo row and membership version are read from the same transactional snapshot during concurrent updates. #​11400
• Route Redis membership and directory operations to the primary. Primary routing provides current membership and registration state and avoids replica write failures during reconnection. #​11484
• Keep elected Kubernetes agents monitoring. Membership updates preserve the elected agent's pod watch and dead-silo pod cleanup; election handoffs and shutdown coordinate pending monitor wakeups. #​11600
• Update KubernetesClient to 21.0.8, with a supported dependency range of [21.0.8,22.0.0). #​11503
• Align consistent-ring routing with published ownership ranges, including unsigned hash positions, range endpoints, and wraparound. #​11598

Client startup and storage cancellation

• Honor client startup cancellation throughout initialization. Cancellation reaches gateway-provider initialization, discovery, connection attempts, retry decisions, and initial cluster-manifest retrieval. A cancelled start releases owned startup resources before completing. After successful startup, background refreshes follow the client's lifetime token. Existing gateway providers remain supported through default implementations of the new cancellation-aware methods. #​11605
• Forward storage cancellation tokens through ADO.NET, Azure Blob, Azure Table, Cosmos DB, DynamoDB, and Firestore grain storage operations. #​11410, #​11411, #​11412, #​11413, #​11414, #​11415

Serialization and packaging

• Restore standalone serializer code generation. Projects referencing Microsoft.Orleans.Serialization receive the transitive source-generator and analyzer dependencies needed for [GenerateSerializer] types, fixing runtime CodecNotFoundException failures caused by missing generated codecs. #​11603
• Deserialize empty multidimensional arrays correctly. #​11430
• Enforce Arc buffer bounds and lifetime rules. Fixes cover empty buffers, page-backed slices, disposal checks, consecutive empty pages, and nested reader forks. #​11474
• Tighten type resolution and authorization. Assembly-qualified lookup respects the specified assembly, and generic type definitions require their own authorization while explicit argument denials remain authoritative. #​11456
• Update MessagePack to 3.1.11, incorporating nested-array, regex backtracking, and LZ4 integer-overflow security fixes. #​11519

... (truncated)

10.4.0

This release strengthens cluster membership consistency, extends cooperative cancellation across framework APIs, and gives streaming applications more control over subscription start positions and dequeue sizes. This release also fixes SQLite persistence under contention, modernizes request-latency metrics, improves serialization and source-generation efficiency, and expands the preview journaling and Durable Jobs capabilities.

Compatibility and upgrade notes

[!IMPORTANT]
Review the SQLite query updates and metric schema changes when upgrading.

• Refresh existing SQLite persistence queries (#​11354). SQLite writes now use a single implicit transaction, preserving atomicity and durable clears when pooled connections encounter writer contention. Reapply the 10.4.0 SQLite main script and SQLite persistence script to existing databases to install the corrected queries. Both scripts are idempotent and preserve grain state.
• Update request-latency dashboards and exporter configuration (#​11251). Application request latency is now a standard Histogram<double> named orleans-app-requests-latency, measured in fractional milliseconds. It replaces the separate orleans-app-requests-latency-bucket, -count, and -sum instruments and their duration attribute. Configure histogram boundaries through an OpenTelemetry View and update dashboards, alerts, and recording rules for the exporter representation. Use matching boundaries across instances whose distributions are aggregated.
• Use canonical grain-type metric dimensions (#​11253). orleans-grains now uses grain_type, containing the canonical GrainId.Type identity, in place of type, which contained the CLR implementation name. Activation lifecycle metrics also gain this dimension. Update queries and label filters accordingly.
• Review explicitly numbered RPC parameters (#​11179). RPC parameter [Id] attributes now control serialized argument IDs, and automatic IDs count serialized parameters. Place CancellationToken last. Existing contracts with parameter [Id] attributes or non-last cancellation tokens can change argument wire IDs relative to 10.3.1; coordinate their client/silo contract upgrade or introduce versioned contracts. Method-level [Id] attributes and ordinary unannotated, token-last contracts retain their existing identities and argument numbering.
• Review preview journaling API changes (#​11211, #​11276, #​11319). Journaling and Durable Jobs packages continue as 10.4.0-alpha.1. Custom integrations should adopt the updated durable-state/state-machine contracts and streamed catalog enumeration APIs. Named-provider migration is an opt-in capability for moving Durable Jobs between storage backends; existing default-provider convenience registrations remain available.

Highlights

Runtime, clustering, and reminders

• Consistent membership across providers. Core membership and the Azure Storage, Cosmos DB, DynamoDB, Consul, Redis, Cassandra, Firestore, ADO.NET, and ZooKeeper implementations strengthen atomic updates, canonical snapshots, and heartbeat handling. ZooKeeper preserves consistent snapshot reads under contention, and Firestore preserves atomic membership reads (#​11307, #​11309, #​11317, #​11342, #​11387).
• Cancellation-aware framework contracts. Trailing cancellation tokens flow through runtime, reminder, streaming, transaction, management, and other framework operations. Legacy overloads and stable wire aliases preserve existing implementations and established framework call identities (#​10937).
• More efficient manifest retrieval. Content-addressed retrieval is enabled by default, allowing silos to reuse immutable manifests by canonical SHA-256 hash and repair missing metadata from peers. ClusterManifestOptions.EnableContentAddressedRetrieval selects the retrieval mode (#​11239).
• Stronger lifecycle handling. Shutdown drains admitted connection-establishment and client-observer work; reminder startup and topology reconciliation are better ordered; directory recovery deactivates duplicate activations; and idle migrated activations are collected (#​11263, #​11264, #​11118, #​11030, #​11365).

Streaming

• Choose where subscriptions begin. StreamSubscriptionStartPosition.Latest and EarliestAvailable provide per-subscription control, with a provider default through StreamPullingAgentOptions.InitialSubscriptionStartPosition. EarliestAvailable starts at the oldest retained message for the stream in the local queue cache. An explicit sequence token takes precedence, and Latest remains the default. Enable the new controls after pulling-agent silos have been upgraded (#​10936).
• Better recovery and handoffs. Typed cache-cursor results, provider-encoded offsets, and a reusable checkpoint state machine improve provider recovery. SQS and Azure Queue handoffs release messages, recovery preserves prefetched batches and pooled-buffer ownership, and pulling agents drain detached work during shutdown (#​9714, #​11153, #​11154, #​11062, #​11144, #​11268).
• Tune memory-stream dequeue sizes. Per-provider MemoryStreamCacheOptions.MaxAddCount controls the maximum number of queue records requested per dequeue, allowing applications to balance response size and queue-call overhead. The default remains 100 (#​11238).

Persistence and transactions

• SQLite persistence returns scalar versions and recovers cleanly from writer contention. Relational providers also receive normalized integer reads, corrected PostgreSQL persistence scripts, and Oracle defunct-silo cleanup support (#​11322, #​11354, #​11283, #​11247, #​11246).
• Transaction participant locks honor transaction timeouts, aborted queued lock operations are removed, and recovery cleanup tasks are awaited (#​10457, #​10455, #​11188).
• DynamoDB providers honor configured endpoint credentials, bind token and profile options, and redact tokens in logs. Provider configuration validation and invariant text handling are strengthened across storage and clustering integrations (#​11292, #​11290, #​11288).

Preview journaling and Durable Jobs

• S3 Express One Zone journal storage. Microsoft.Orleans.Journaling.S3 adds append-based write-ahead logs, conditional metadata/checkpoint publication, catalog discovery, and provider metrics. A conditional-rewrite mode supports S3-compatible development environments (#​10161).
• Composable, named journal providers. Grains can compose durable state through the updated APIs, while named providers bind storage, catalogs, and state-manager factories together. Catalog enumeration streams results, and recovery serializes retries while preserving persistence failures (#​11279, #​11275, #​11211, #​11326).
• Controlled Durable Jobs provider migration. Select an active provider for new shards and draining providers for existing jobs, inspect storage before retirement, and retain saved job handles across the cutover. Shard discovery uses start-time ordering, and shutdown drains scheduling and releases running shards. Queue wake-ups and flush signals are coalesced (#​11277, #​11258, #​11301, #​11103, #​11206).

Serialization, code generation, and deployment

• Improved NativeAOT building blocks. Generated C# codecs and copiers use ref-returning UnsafeAccessor methods for private, readonly, and backing fields on supported targets. Grain activation constructors are preserved, non-generic grain references gain native construction, and serialization constructors can initialize existing objects under NativeAOT. Default serializer activation uses a shared construction path with consistent exception propagation (#​11372, #​11383, #​11371, #​11373, #​11390).
• Opt-in serializer Hot Reload. Set <OrleansHotReload>true</OrleansHotReload> to generate stable member/dependency identities and lazily initialize dependencies added to existing serializers and copiers. Supported updates add strongly typed serialized members to classes and records; restart to refresh manifests when adding new polymorphic or collection-element types (#​10932).
• Correct compound-aliased array resolution. Polymorphic arrays and arrays nested in closed generic types deserialize correctly, with the existing writer bytes preserved (#​11392).
• Less repeated work and copying. The generator caches library types per compilation, serialization-attribute analysis skips generated code, contiguous JSON payloads are read directly, and buffered tail reads avoid repeated page traversal. Metadata discovery also supports single-file deployment (#​11363, #​11384, #​11270, #​11209, #​11054).

Observability and developer tooling

• Request latency uses native histogram aggregation, while canonical, cached grain-type dimensions make activation populations and lifecycle events easier to correlate (#​11251, #​11253).
  ... (truncated)

Commits viewable in compare view.

Updated Microsoft.Orleans.Server from 10.3.1 to 10.4.2.

Release notes

Sourced from Microsoft.Orleans.Server's releases.

10.4.2

What's Changed

• [10.x] fix(azure): tolerate legacy membership writers during rolling upgrades by @​ReubenBond in [10.x] fix(azure): tolerate legacy membership writers during rolling upgrades dotnet/orleans#11624

Full Changelog: dotnet/orleans@v10.4.1...v10.4.2

10.4.1

Orleans 10.4.1

Orleans 10.4.1 improves cluster membership, grain scheduling, graceful shutdown, client startup, serialization, and provider reliability. It also restores standalone serializer code generation, fixes ADO.NET streaming message loss, and improves dashboard statistics and navigation.

Compatibility and upgrade notes

[!IMPORTANT]
Public API compatibility: SiloAddress.GetUniformHashCodes(int) now returns ImmutableArray<uint> instead of uint[]. This is a source- and binary-breaking change for code which calls this method. Rebuild affected applications and libraries against 10.4.1. Use ToArray() when a mutable array is required, or Sort() for an immutable sorted result. See #​11425.

• Distributed grain directory: Preserve the cluster's existing partition count and ordered partition boundaries throughout an upgrade. The new GrainDirectoryOptions.GetPartitionBoundaries callback supports configuring the mapping used by an existing cluster, including the older sorted-hash mapping. Partition defaults remain unchanged. See #​11425.
• Management deactivation: Begin using the new IManagementGrain.DeactivateGrain operation after all active silos have been upgraded to support it. See #​11483.
• Custom consistent-ring configurations: Clusters using UseVirtualBucketsConsistentRing = false receive corrected request routing. During a rolling upgrade, older callers can still route reminder operations to the wrong owner. Persisted reminders are reconciled on the owner's next successful refresh, which defaults to five minutes; early occurrences can be missed before reconciliation. Custom grain services using this ring need application-specific upgrade handling. The default virtual-buckets provider retains its existing behavior. See #​11598.

What's changed

The release preparation PR, #​11566, bundles multiple backports. Those changes are expanded below, with links to their original PRs alongside the subsequent 10.x fixes.

Grain execution and graceful shutdown

• Preserve read-only scheduling isolation. Request admission accounts for every active request's interleaving policy, keeping writable requests isolated from active [ReadOnly] requests and correctly handling mixed MayInterleave workloads. #​10809
• Make grain timer scheduling consistent. Timer ticks coalesce while a callback is queued or running, delayed wakeups are preserved, and Change calls made during that interval apply the latest schedule after the callback completes. #​11464
• Carry queued requests with migrating activations. Graceful silo retirement and explicit activation migration transfer eligible queued requests through normal destination admission. Best-effort caller retargeting reduces SiloUnavailableException failures when a retiring silo departs, while preserving callers' original deadlines. Applications should continue to account for ambiguous outcomes when an acceptance acknowledgement is lost. #​11460
• Run observer-level incoming call filters. Filters implemented by observer objects participate in incoming call processing. #​11507
• Safely deactivate existing grain instances. IManagementGrain.DeactivateGrain(GrainId, CancellationToken) locates existing activations directly and waits for their deactivation callbacks. This supports lifecycle operations on grains whose persisted state cannot currently be read. #​11483

Cluster membership, directories, and Kubernetes

• Preserve the default grain directory when registering a named distributed directory. Grains continue to use their configured directory, fixing duplicate activations caused by the registration interaction. #​11553
• Keep membership snapshots canonical and retain departed-silo knowledge. Local startup lifecycle tracking remains separate from published membership rows. Retained death information lets client routing and durable-job adoption account for silos after their rows are pruned. Snapshot decoding also supports older silos which omit the new history field. #​11593
• Improve membership health monitoring. Health monitors handle shutdown cancellation cleanly, and joining silos reevaluate stale peers. #​11421, #​11431
• Upgrade legacy Azure Table membership data safely. Existing legacy tables can be read with their versions and ETags intact. The first successful membership write atomically establishes the snapshot boundary markers, and reads retry when they overlap marker installation. #​11596
• Read Firestore membership rows atomically. The silo row and membership version are read from the same transactional snapshot during concurrent updates. #​11400
• Route Redis membership and directory operations to the primary. Primary routing provides current membership and registration state and avoids replica write failures during reconnection. #​11484
• Keep elected Kubernetes agents monitoring. Membership updates preserve the elected agent's pod watch and dead-silo pod cleanup; election handoffs and shutdown coordinate pending monitor wakeups. #​11600
• Update KubernetesClient to 21.0.8, with a supported dependency range of [21.0.8,22.0.0). #​11503
• Align consistent-ring routing with published ownership ranges, including unsigned hash positions, range endpoints, and wraparound. #​11598

Client startup and storage cancellation

• Honor client startup cancellation throughout initialization. Cancellation reaches gateway-provider initialization, discovery, connection attempts, retry decisions, and initial cluster-manifest retrieval. A cancelled start releases owned startup resources before completing. After successful startup, background refreshes follow the client's lifetime token. Existing gateway providers remain supported through default implementations of the new cancellation-aware methods. #​11605
• Forward storage cancellation tokens through ADO.NET, Azure Blob, Azure Table, Cosmos DB, DynamoDB, and Firestore grain storage operations. #​11410, #​11411, #​11412, #​11413, #​11414, #​11415

Serialization and packaging

• Restore standalone serializer code generation. Projects referencing Microsoft.Orleans.Serialization receive the transitive source-generator and analyzer dependencies needed for [GenerateSerializer] types, fixing runtime CodecNotFoundException failures caused by missing generated codecs. #​11603
• Deserialize empty multidimensional arrays correctly. #​11430
• Enforce Arc buffer bounds and lifetime rules. Fixes cover empty buffers, page-backed slices, disposal checks, consecutive empty pages, and nested reader forks. #​11474
• Tighten type resolution and authorization. Assembly-qualified lookup respects the specified assembly, and generic type definitions require their own authorization while explicit argument denials remain authoritative. #​11456
• Update MessagePack to 3.1.11, incorporating nested-array, regex backtracking, and LZ4 integer-overflow security fixes. #​11519

... (truncated)

10.4.0

This release strengthens cluster membership consistency, extends cooperative cancellation across framework APIs, and gives streaming applications more control over subscription start positions and dequeue sizes. This release also fixes SQLite persistence under contention, modernizes request-latency metrics, improves serialization and source-generation efficiency, and expands the preview journaling and Durable Jobs capabilities.

Compatibility and upgrade notes

[!IMPORTANT]
Review the SQLite query updates and metric schema changes when upgrading.

• Refresh existing SQLite persistence queries (#​11354). SQLite writes now use a single implicit transaction, preserving atomicity and durable clears when pooled connections encounter writer contention. Reapply the 10.4.0 SQLite main script and SQLite persistence script to existing databases to install the corrected queries. Both scripts are idempotent and preserve grain state.
• Update request-latency dashboards and exporter configuration (#​11251). Application request latency is now a standard Histogram<double> named orleans-app-requests-latency, measured in fractional milliseconds. It replaces the separate orleans-app-requests-latency-bucket, -count, and -sum instruments and their duration attribute. Configure histogram boundaries through an OpenTelemetry View and update dashboards, alerts, and recording rules for the exporter representation. Use matching boundaries across instances whose distributions are aggregated.
• Use canonical grain-type metric dimensions (#​11253). orleans-grains now uses grain_type, containing the canonical GrainId.Type identity, in place of type, which contained the CLR implementation name. Activation lifecycle metrics also gain this dimension. Update queries and label filters accordingly.
• Review explicitly numbered RPC parameters (#​11179). RPC parameter [Id] attributes now control serialized argument IDs, and automatic IDs count serialized parameters. Place CancellationToken last. Existing contracts with parameter [Id] attributes or non-last cancellation tokens can change argument wire IDs relative to 10.3.1; coordinate their client/silo contract upgrade or introduce versioned contracts. Method-level [Id] attributes and ordinary unannotated, token-last contracts retain their existing identities and argument numbering.
• Review preview journaling API changes (#​11211, #​11276, #​11319). Journaling and Durable Jobs packages continue as 10.4.0-alpha.1. Custom integrations should adopt the updated durable-state/state-machine contracts and streamed catalog enumeration APIs. Named-provider migration is an opt-in capability for moving Durable Jobs between storage backends; existing default-provider convenience registrations remain available.

Highlights

Runtime, clustering, and reminders

• Consistent membership across providers. Core membership and the Azure Storage, Cosmos DB, DynamoDB, Consul, Redis, Cassandra, Firestore, ADO.NET, and ZooKeeper implementations strengthen atomic updates, canonical snapshots, and heartbeat handling. ZooKeeper preserves consistent snapshot reads under contention, and Firestore preserves atomic membership reads (#​11307, #​11309, #​11317, #​11342, #​11387).
• Cancellation-aware framework contracts. Trailing cancellation tokens flow through runtime, reminder, streaming, transaction, management, and other framework operations. Legacy overloads and stable wire aliases preserve existing implementations and established framework call identities (#​10937).
• More efficient manifest retrieval. Content-addressed retrieval is enabled by default, allowing silos to reuse immutable manifests by canonical SHA-256 hash and repair missing metadata from peers. ClusterManifestOptions.EnableContentAddressedRetrieval selects the retrieval mode (#​11239).
• Stronger lifecycle handling. Shutdown drains admitted connection-establishment and client-observer work; reminder startup and topology reconciliation are better ordered; directory recovery deactivates duplicate activations; and idle migrated activations are collected (#​11263, #​11264, #​11118, #​11030, #​11365).

Streaming

• Choose where subscriptions begin. StreamSubscriptionStartPosition.Latest and EarliestAvailable provide per-subscription control, with a provider default through StreamPullingAgentOptions.InitialSubscriptionStartPosition. EarliestAvailable starts at the oldest retained message for the stream in the local queue cache. An explicit sequence token takes precedence, and Latest remains the default. Enable the new controls after pulling-agent silos have been upgraded (#​10936).
• Better recovery and handoffs. Typed cache-cursor results, provider-encoded offsets, and a reusable checkpoint state machine improve provider recovery. SQS and Azure Queue handoffs release messages, recovery preserves prefetched batches and pooled-buffer ownership, and pulling agents drain detached work during shutdown (#​9714, #​11153, #​11154, #​11062, #​11144, #​11268).
• Tune memory-stream dequeue sizes. Per-provider MemoryStreamCacheOptions.MaxAddCount controls the maximum number of queue records requested per dequeue, allowing applications to balance response size and queue-call overhead. The default remains 100 (#​11238).

Persistence and transactions

• SQLite persistence returns scalar versions and recovers cleanly from writer contention. Relational providers also receive normalized integer reads, corrected PostgreSQL persistence scripts, and Oracle defunct-silo cleanup support (#​11322, #​11354, #​11283, #​11247, #​11246).
• Transaction participant locks honor transaction timeouts, aborted queued lock operations are removed, and recovery cleanup tasks are awaited (#​10457, #​10455, #​11188).
• DynamoDB providers honor configured endpoint credentials, bind token and profile options, and redact tokens in logs. Provider configuration validation and invariant text handling are strengthened across storage and clustering integrations (#​11292, #​11290, #​11288).

Preview journaling and Durable Jobs

• S3 Express One Zone journal storage. Microsoft.Orleans.Journaling.S3 adds append-based write-ahead logs, conditional metadata/checkpoint publication, catalog discovery, and provider metrics. A conditional-rewrite mode supports S3-compatible development environments (#​10161).
• Composable, named journal providers. Grains can compose durable state through the updated APIs, while named providers bind storage, catalogs, and state-manager factories together. Catalog enumeration streams results, and recovery serializes retries while preserving persistence failures (#​11279, #​11275, #​11211, #​11326).
• Controlled Durable Jobs provider migration. Select an active provider for new shards and draining providers for existing jobs, inspect storage before retirement, and retain saved job handles across the cutover. Shard discovery uses start-time ordering, and shutdown drains scheduling and releases running shards. Queue wake-ups and flush signals are coalesced (#​11277, #​11258, #​11301, #​11103, #​11206).

Serialization, code generation, and deployment

• Improved NativeAOT building blocks. Generated C# codecs and copiers use ref-returning UnsafeAccessor methods for private, readonly, and backing fields on supported targets. Grain activation constructors are preserved, non-generic grain references gain native construction, and serialization constructors can initialize existing objects under NativeAOT. Default serializer activation uses a shared construction path with consistent exception propagation (#​11372, #​11383, #​11371, #​11373, #​11390).
• Opt-in serializer Hot Reload. Set <OrleansHotReload>true</OrleansHotReload> to generate stable member/dependency identities and lazily initialize dependencies added to existing serializers and copiers. Supported updates add strongly typed serialized members to classes and records; restart to refresh manifests when adding new polymorphic or collection-element types (#​10932).
• Correct compound-aliased array resolution. Polymorphic arrays and arrays nested in closed generic types deserialize correctly, with the existing writer bytes preserved (#​11392).
• Less repeated work and copying. The generator caches library types per compilation, serialization-attribute analysis skips generated code, contiguous JSON payloads are read directly, and buffered tail reads avoid repeated page traversal. Metadata discovery also supports single-file deployment (#​11363, #​11384, #​11270, #​11209, #​11054).

Observability and developer tooling

• Request latency uses native histogram aggregation, while canonical, cached grain-type dimensions make activation populations and lifecycle events easier to correlate (#​11251, #​11253).
  ... (truncated)

Commits viewable in compare view.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore <dependency name> major version will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself)
• @dependabot ignore <dependency name> minor version will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself)
• @dependabot ignore <dependency name> will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself)
• @dependabot unignore <dependency name> will remove all of the ignore conditions of the specified dependency
• @dependabot unignore <dependency name> <ignore condition> will remove the ignore condition of the specified dependency and ignore conditions