Bump OpenTelemetry.Exporter.OpenTelemetryProtocol from 1.17.0 to 1.18.0 - #87
Closed
dependabot[bot] wants to merge 643 commits into
Closed
Conversation
added 30 commits
June 10, 2026 02:29
- Workflows 1.9.0 changed AgentFileSkillScriptRunner delegate signature from AIFunctionArguments to JsonElement? with a new IServiceProvider? param; updated RunSkillScriptAsync to match and parse JSON properties
- Worksheet and Slide can be null per the OpenXml SDK's nullability annotations; existing null guards already handled the case, just needed ?. to flow null through without a compiler warning
- ExpandSessionPaths calls Path.GetFullPath, which prepends the CWD to
any non-rooted string; applied to multi-line agent Instructions this
corrupts the entire text and leaves ~/ references unexpanded, causing
agents to write session artifacts (e.g. brief.json) to the wrong location
- New ExpandTextTokens replaces {session_id}, {project_slug}, and ~/
inline without path normalisation; InterpolateSessionId now uses it for
Instructions while all file-path fields continue to use ExpandSessionPaths
- Single-turn blowups reached 186k tokens because MaxSingleTurnInputTokens (200k) and InTurnToolWindow (default 20) were too permissive; reducing them forces earlier compaction and tombstones stale tool results more aggressively - Developer MaxTurnAge 8→5 and MaxInTurnContextTokens 50k→30k directly cut the cross-turn history that drove per-turn token explosions - Verifier lacked a pattern for build-unit misattribution; agents patched the wrong artefact when a parent build unit pulled in a child file via glob — new pattern d surfaces this with a specific hypothesis
- MaxSingleTurnInputTokens 120k was below the cost of a routine heavy turn, causing compaction to fire on nearly every Developer/Verifier/Tester turn; raised to 200k so cumulative CutoverAt remains the primary guard - KeepRecentTurns 12→8 so each compaction drops more retained turns and the post-compaction context is meaningfully smaller, breaking the immediate re-trigger loop - InTurnToolWindow 8→5 and MaxToolResultTokens 8k→6k so the tool-result slice sent to the LLM is tighter, compounding the retained-turn reduction - CommandRun evidence already stored up to 4096 chars of output but ContextRebuilder never emitted it; failed commands now include a 400-char snippet so agents resuming after compaction see the actual error text, not just an exit code
- Full-width bordered panels for turns with no text content were visually noisy, consuming screen space to show only a tool-call count - Tool-only turns now render as a compact dim inline summary, keeping the same metadata (agent name, turn index, elapsed, token usage)
- The standard Serilog Console sink writes directly to Console.Out, bypassing Spectre.Console's live-display cursor management; this caused WRN/INF lines to land on the same terminal row as spinner text or compact turn summaries - AnsiConsoleSink routes all formatted log output through AnsiConsole.MarkupLine so Serilog respects the Status spinner and never corrupts the live display layout - VSCode mode keeps the original Console sink with stderr routing so stdout remains a clean JSON stream for the webview bridge
- Verifier was executing verify_command whenever the change log showed it hadn't run yet, including during Planning/BriefReview before the Developer existed any files — causing spurious INCONSISTENCY findings that triggered ConflictingEvidence reinstruct and blocked the handoff to Implementation - Step 4 now skips the shell_run unless SignificantChanges shows at least one files_to_change entry has been written, i.e. implementation has actually started
- PlannerCritic used sub_agent_explore to find "affected files" on every turn, but for new projects this returns a theoretical list of files to create, not files that exist and need modification — causing it to block indefinitely on files the Developer was supposed to create - Critic step 2 now first checks existence via sub_agent_locate; skips exploration entirely if no files_to_change exist yet (greenfield) - Verifier step 2b adds an early-exit guard: if no files_to_change appear in SignificantChanges, skip all inconsistency checks — pre-implementation change log state is not meaningful evidence of a real inconsistency
- ApplyWithManifest appends a [Context Manifest] listing active and superseded tool results so the model knows what it can still see without re-reading full history at the recency end of the prompt - ContextAssembler repeats a brief task reminder at the end of any assembled context exceeding 2 000 chars, sandwiching the objective at both the primacy and recency positions - Tombstones now include the evicted tool label and a 300-char content preview so the model can judge whether to re-read without issuing a blind full-file read
- ApplyWithManifest previously called Apply then rebuilt callLabels with a second scan of the original context; ApplyCore does one pass and returns the map + evicted flag, eliminating the redundant work - extract task-reminder thresholds into named constants in ContextAssembler to make the intent readable without inline comments - add tests for the all-evicted manifest path and the disabled-budget same-reference fast path
- Task Reminder and Context Manifest are non-obvious behavioral features that affect how agents perceive their own context; worth calling out so contributors know to preserve the primacy+recency sandwich invariant - Tombstone format changed (now includes label + preview) — example in the doc prevents stale assumptions about the old single-line format - Add ContextAssembler and ToolResultWindowTrimmer to Where-to-look table
- Tombstone example was the old one-liner; now shows the enriched format with tool label and content preview that ships in this branch - Context Manifest (appended when evictions occur) was not documented - Task Reminder step was missing from both pipeline overview and diagram
- Session context was injected at position 1, burying it mid-history where models attend least; moved to recency boundary so it is read last - TrimToWindow could drain the final message pair: the assistant removal only checked start < list.Count, not start + 1 < list.Count, allowing both messages to be removed in a single loop iteration - Pipeline Knowledge was re-appended every turn regardless of whether identical content already existed in history, compounding context growth
- Roslyn on Windows sees two resolution paths for ContextAddCommand when both fuseraft.Cli.Commands and fuseraft.Cli.Commands.Context are imported, triggering an ambiguous reference error; type aliases make each name resolve to exactly one type on all platforms
- Bar chart at top of ctx_viz.html shows input and output tokens per turn with toggle buttons to show/hide each dataset independently - Events.jsonl is now loaded alongside ctx_snapshots.jsonl; validation_fail and tool_blocked events render as annotated vertical lines on both charts - Context assembly details (chars, tool count, assembly time) from events.jsonl surface in bar and line chart tooltips when present - Only four event types are loaded (turn_end, validation_fail, tool_blocked, context_assembly) to keep embedded HTML size reasonable
- Inline string literals scattered across 18 files created silent breakage risk: a typo in any emitter or consumer would compile fine but produce events that no hook, filter, or viz would ever match - const string fields work as switch patterns (unlike static readonly), so EventLogViewer's switch expressions adopt them without ceremony
- Eliminates 47+ raw string literals across Cli, Infrastructure, and Orchestration layers — typos caused silent data gaps in events.jsonl with no compile-time signal - Extends EventTypes with all missing constants so every event string has a single authoritative definition
- Aligns naming with context_budget_warn so the distinction is obvious: budget = token accumulation, window = message-count cap
- Both events signaled the same condition from different layers; a source field in the payload now distinguishes graph_orchestrator, keyword_strategy, and state_machine_strategy without a separate type
Defines agent lifecycle (AgentStart/End/Error/Timeout), tool completion (ToolResult/Error/Timeout), retry (RetryScheduled/Attempt/Exhausted), termination (TerminationSatisfied/MaxTurnsExceeded), parallel branches (ParallelBranchStart/End/Error), HITL outcomes (HitlApproved/Rejected/ Resolved), model invocation, selection strategy, knowledge retrieval, artifact lifecycle, and checkpointing/replay constants. Wires emit calls in AgentOrchestrator, GraphOrchestrator, ChangeTracker (CapturingMiddleware), and SessionRunner. Adds four SessionRunner tests covering CancellationRequested, MaxTurnsExceeded, and HitlResolved (agent-blocked and validator-stuck paths) using hook-based synchronization for fire-and-forget emits. false
- Checkpoint lifecycle (created/loaded/resume_started/resume_completed, event_replay_start/complete) needed emission sites; wired in RunCommand and SessionRunner where the emitter is already in scope - Model invocation events (model_call/response/error/timeout) wired in AgentFactory middleware chain alongside existing inner_call_context; model_timeout also added to GraphOrchestrator streaming catch sites so SSE idle timeouts surface at the model level, not just turn/agent - cancellation_observed distinguishes clean inter-turn stops from mid-turn OperationCanceledException (cancellation_requested); wired at the proactive IsCancellationRequested check in SessionRunner - JsonSessionStore gains OnCorruptionDetected callback for event_corruption_detected; avoids circular dependency by using a Func delegate instead of a direct EventEmitter reference - EventEmitter docblock updated to reference EventTypes instead of listing a stale hand-picked subset - docs/sessions.md gains an orchestration event types reference table covering all wired event groups
- Eliminates typo-induced silent failures by centralising the canonical strings in one place; a rename now only touches OrchestratorTypes.cs - Covers ValidateConfigCommand, WorkflowDiagramGenerator, OrchestratorBuilder, StrategyFactory, and KeywordSelectionStrategy
- Eliminates typo-induced silent failures; a mode rename now only touches CompactionModes.cs - CompactionConfig.Mode default left as a bare string to avoid introducing a Core.Models → Orchestration layer dependency
- Mirrors the OrchestratorTypes/CompactionModes pattern to eliminate scattered inline literals that could silently diverge - Capitalizes "orchestrator" → "Orchestrator" for consistency with all other reserved names (System, Human, Assistant, Verifier, Unknown)
- Eliminates scattered inline string literals for validator names, matching the pattern established by AgentNames and CompactionModes - Replaces ToLowerInvariant switch in GraphOrchestrator with OrdinalIgnoreCase if/else chain for consistency with StrategyFactory
- Replaces raw "assistant"/"user" string literals with MessageRole constants to eliminate scatter across orchestrators and CLI - Moves CompactionReason out of CompactionCoordinator into its own file, consistent with CompactionModes and similar constants classes
An audit of the actual agent-framework source (checked against Microsoft.Agents.AI.Workflows 1.16.0, the version we pin) found three claims in the "why we don't use X" rationale no longer hold, and two framework features we do use weren't documented at all. - Handoff orchestration: autonomous mode now injects a continuation message and re-invokes the agent when the handoff tool isn't called — the "no correction-injection loop, blocks for human input" framing is stale, and autonomous mode graduated out of experimental. The real remaining gap is generic vs. typed/validator-driven corrections, and the shared-history/ContextWindow-filtering incompatibility. - Concurrent orchestration: branches don't share one mutable AgentContext (that's a fuseraft type, not a MAF one, and a category error to begin with) — each agent gets its own session-isolated executor. The actual blocker is retry semantics, which stands on its own without the shared-history argument. - GraphOrchestrator cycles: WorkflowBuilder.Validate() never checked for cycles, and the framework's own GroupChatWorkflowBuilder/HandoffWorkflowBuilder build cyclic graphs by design. The real reason for phase-restart is validator gating at phase boundaries, not a framework limitation. - Added missing "what we use" entries: Microsoft.Agents.AI.Compaction (ToolResultCompactionStrategy/CompactionProvider, used in AgentContextCompactionFilters) and AgentSkillsProvider/ AgentSkillsProviderBuilder (MAF's own Skills feature, not a fuseraft-built system despite reading like one). - Softened the GroupChatWorkflowBuilder rejection: a manager subclass can hold private state via the framework's checkpoint hooks; what's actually forced is that GroupChatHost always passes full history into the manager's own decision calls, which is what Magentic's two-history invariant is actually incompatible with. None of this changes fuseraft-cli's own code or its architectural decisions — the underlying reasons (typed correction pipeline, per-agent ContextWindow filtering, two-history Magentic invariant) still hold. This only corrects the framework-behavior claims used to justify them.
- Documents why TruncationCompactionStrategy/SummarizationCompactionStrategy aren't used for session-level compaction: they operate on ChatMessage, not fuseraft's AgentMessage/AgentContext.History, and neither expresses the routing-signal-pinning or grounding invariants ConversationCompactor already implements
- Free-form turns only ever appended the final text to ctx.History, dropping every FunctionCallContent/FunctionResultContent from the turn — the model had no record of its own tool calls past the current turn. Now the full transcript (assistant calls, tool-role results, final text) is captured via ChatResponseExtensions.AddMessages, matching what TrimHistory/compaction filters/session snapshotting already expected but never received. - A cancelled or unrecoverable streaming error mid-/execute cleared the entire remaining plan queue with HaltedAt never set, so /resume and /recover had nothing to act on — infra hiccups and deliberate Ctrl+C got the same irreversible treatment as a designed stop. Those failures now route through the same halted-state transition a step verify-failure already gets. - Free-form/plan-capture turns silently hit ctx.Client's tool-iteration cap with no signal — the model gets cut off mid tool-loop and returns a rushed answer that looks like a normal complete response. Step mode already warned on this; free-form turns now do too.
- Live-tested evidence shows grok-4.3 tolerates ~58 tools without degradation, but every registered tool still adds its schema to every request, so the token/latency cost is real regardless - Split the always-on FileSystem/Shell/Git surface into a curated core (read, edit, search, status, commit — ~37 tools total) and an opt-in "Extended" plugin (delete/copy/move, remote git, background shell jobs — ~24 tools), enabled the same way as Http/Changes/etc. via --plugins Extended - /explore and /locate keep full read-tool access regardless of whether Extended is enabled, since SubAgentPlugin's explorerTools is built from the unfiltered FileSystem/Shell/Git lists
- TodoPlugin was always constructed fresh in ReplCommand.ExecuteAsync and ReplSessionSnapshot never captured its state, so after --resume the restored chat history could show a populated checklist while todo_read returned [EMPTY] — the one tool whose entire point is durable state was the one piece of REPL state --resume dropped - Add TodoPlugin.Restore() to set state directly from a prior snapshot without re-running Write's JSON parsing/validation - Thread todo items through ReplSessionSnapshot.Capture/SaveAsync and restore them in the --resume path, mirroring how CurrentPlan/ ExecutionQueue/HaltedAt are already persisted
- MagenticWorkflowBuilder ships in the pinned MAF 1.16.0 and is purpose-built for Magentic-One orchestration, but its manager reasoning calls thread the same shared taskContext.ChatHistory that participants see (verified against Specialized/Magentic/ MagenticOrchestrator.cs and MagenticManager.cs) — same defect as GroupChatWorkflowBuilder, just Magentic-specific, so it can't express our manager-never-sees-raw-history invariant (§6.2)
- A turn's stream can end without throwing right after a tool_use chunk, before its tool_result ever arrives (dropped connection mid tool-round). The unresolved call still got merged into ctx.History, and providers reject any request whose history ends with a tool_use lacking a matching tool_result — permanently 400ing every later turn in that session with no self-repair. - RepairDanglingToolCalls scans history for orphaned FunctionCallContent and appends a synthetic tool_result so the transcript stays valid. Runs after merging a turn's raw updates, and again after restoring a --resume snapshot so a session poisoned by an earlier (unpatched) run heals instead of immediately failing again.
- CmdForkAsync built its own ReplSessionSnapshot.Capture(...) call instead of reusing ReplTurn.SaveSnapshotAsync, and never passed todoItems — a forked session (or --resume of one) silently lost every todo_write the model had made, even though 7124f5c already fixed the same gap for the ordinary auto-save/--resume path - Add ReplForkTodoPersistenceTests: the regression case (todo items survive a fork), plus two edge cases (empty todo list stays null, no TodoPlugin instance doesn't throw) Claude-Session: https://claude.ai/code/session_016inyJ2gaJQiyMZsWtjva3X
- OnCancelKeyPress only suppressed the default SIGINT action while a turn was running or in JSON mode; idle at the prompt it did nothing, so .NET's default action terminated the process outright (exit code 130) with no "Session ended.", no memory extraction, and no snapshot of the in-progress line. - Ctrl+C is consumed by the terminal's SIGINT/ISIG machinery before it ever reaches Console.ReadKey, so CancelKeyPress is the only place that ever observes an idle-prompt Ctrl+C — ReplLineReader's own ConsoleKey.C case was dead code in every terminal actually tested. - Every other REPL (and this app's own mid-turn behavior) treats Ctrl+C as "abandon the current line/operation," not "quit," so match that instead of leaving the two cases inconsistent.
- Cursor movement, backspace, and delete moved one UTF-16 code unit at a time, so they could land between the two halves of a surrogate pair (any emoji outside the BMP, e.g. 🚀) and split it apart — editing near one silently mangled it into replacement characters. Added StepBack/StepForward helpers so every move/delete steps a full character. - The hardcoded SlashCommands tab-completion list had drifted from the dispatcher: /run, /models, /reasoning, and /snapshot were all real, /help-documented commands that Tab could never complete to.
- /paste help said "type EOF to finish" but the actual prompt asks for .done or Ctrl+D; found while manually driving the REPL through tmux - code block panels always expanded to full terminal width regardless of content, dwarfing short snippets next to snugly-fit tables; now sized to content (still wraps/expands correctly for long lines) and shows the fenced language tag in the panel header - /sessions had no column headers and truncated model ids at 22 chars, chopping common longer ids like grok-4-1-fast-reasoning
--- updated-dependencies: - dependency-name: Microsoft.Agents.AI.OpenAI dependency-version: 1.17.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
--- updated-dependencies: - dependency-name: ModelContextProtocol dependency-version: 2.1.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
--- updated-dependencies: - dependency-name: OllamaSharp dependency-version: 5.4.30 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
….0 (#75) --- updated-dependencies: - dependency-name: OpenTelemetry.Exporter.OpenTelemetryProtocol dependency-version: 1.17.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
--- updated-dependencies: - dependency-name: OpenTelemetry.Instrumentation.Runtime dependency-version: 1.17.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
--- updated-dependencies: - dependency-name: SQLitePCLRaw.bundle_e_sqlite3 dependency-version: 3.0.5 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
--- updated-dependencies: - dependency-name: Microsoft.Agents.AI.Workflows dependency-version: 1.17.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Scott Stauffer <scott@fuseraft.com>
--- updated-dependencies: - dependency-name: OpenTelemetry.Instrumentation.Http dependency-version: 1.17.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Scott Stauffer <scott@fuseraft.com>
UserConfigStoreLegacyKeyFileTests, FuseraftPathsHomeOverrideTests, and ReplForkTodoPersistenceTests each set the process-wide FUSERAFT_HOME env var. xUnit gives each test class its own collection by default and runs collections in parallel, so these three classes could race: FuseraftPathsHomeOverrideTests briefly clearing FUSERAFT_HOME to null could interleave with another class expecting its own override, making it read the real ~/.fuseraft/config instead of its temp dir. Group them into one xUnit collection so their tests run sequentially.
Every /exit path (slash command, bare "exit"/"quit", Ctrl+D) now prints the exact `fuseraft --resume <sessionId>` command after "Session ended." so the session ID doesn't have to be dug out of /sessions or logs. Claude-Session: https://claude.ai/code/session_01U9Ytg59a8MuHrTkJ38wQ1E
ReasoningEffort is a pure pass-through — injected verbatim as
"reasoning": {"effort": "..."} and never interpreted by fuseraft — but
/model, /reasoning, and `validate` all rejected anything outside a
hardcoded none/low/medium/high list, blocking newer provider tiers like
"minimal", "xhigh", and "max".
Replace the fixed allow-list with a light sanity check (reject only
values with embedded whitespace, which are unambiguously malformed)
so new provider-specific values work without a CLI update. Update help
text and docs accordingly.
Claude-Session: https://claude.ai/code/session_01U9Ytg59a8MuHrTkJ38wQ1E
--- updated-dependencies: - dependency-name: OpenTelemetry.Exporter.OpenTelemetryProtocol dependency-version: 1.18.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
.NET
Contributor
Author
|
OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting If you change your mind, just re-open this PR and I'll resolve any conflicts on it. |
dependabot
Bot
deleted the
dependabot/nuget/src/OpenTelemetry.Exporter.OpenTelemetryProtocol-1.18.0
branch
August 25, 2026 03:53
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Updated OpenTelemetry.Exporter.OpenTelemetryProtocol from 1.17.0 to 1.18.0.
Release notes
Sourced from OpenTelemetry.Exporter.OpenTelemetryProtocol's releases.
1.18.0
For highlights and announcements pertaining to this release see: Release Notes > 1.18.0.
The following changes are from the previous release 1.17.0.
NuGet: OpenTelemetry v1.18.0
Fixed self-diagnostics log lines being silently dropped when an event message or parameter contained enough 3-byte UTF-8 characters to overflow the internal buffer estimate. Such content is now truncated.
(#7543)
Fixed activity creation throwing when multiple tracer providers return a sampler attribute with the same key.
(#7558)
Added the
otel.sdk.processor.log.processedSDK self-observability metric.(#7486)
Added the
otel.sdk.processor.span.processedSDK self-observability metric.(#7598)
BatchActivityExportProcessorandSimpleActivityExportProcessorno longer forward spans to the exporter onceShutdownhas been called, andBatchActivityExportProcessor.Shutdownnow waits for in-flightOnEndcalls to finish enqueueing before flushing.(#7598)
Fix logger, meter and tracer providers leaking background threads if an exception is thrown by their constructor after resource creation.
(#7615)
CircularBufferBuckets.Copyoptimized to use bulk array copies.(#7670)
Restored configured
MaxScaleafter delta exponential histogram collection.(#7671)
See CHANGELOG for details.
NuGet: OpenTelemetry.Api v1.18.0
RecordExceptionis called on a span that is not recorded.(#7669)
See CHANGELOG for details.
NuGet: OpenTelemetry.Api.ProviderBuilderExtensions v1.18.0
No notable changes.
See CHANGELOG for details.
NuGet: OpenTelemetry.Exporter.Console v1.18.0
IEnumerable<KeyValuePair<string, object?>>). These attributes will be serialized as JSON objects.(#7015)
... (truncated)
1.18.0-rc.1
The following changes are from the previous release 1.17.0.
NuGet: OpenTelemetry v1.18.0-rc.1
Fixed self-diagnostics log lines being silently dropped when an event message or parameter contained enough 3-byte UTF-8 characters to overflow the internal buffer estimate. Such content is now truncated.
(#7543)
Fixed activity creation throwing when multiple tracer providers return a sampler attribute with the same key.
(#7558)
Added the
otel.sdk.processor.log.processedSDK self-observability metric.(#7486)
Added the
otel.sdk.processor.span.processedSDK self-observability metric.(#7598)
BatchActivityExportProcessorandSimpleActivityExportProcessorno longer forward spans to the exporter onceShutdownhas been called, andBatchActivityExportProcessor.Shutdownnow waits for in-flightOnEndcalls to finish enqueueing before flushing.(#7598)
Fix logger, meter and tracer providers leaking background threads if an exception is thrown by their constructor after resource creation.
(#7615)
CircularBufferBuckets.Copyoptimized to use bulk array copies.(#7670)
Restored configured
MaxScaleafter delta exponential histogram collection.(#7671)
See CHANGELOG for details.
NuGet: OpenTelemetry.Api v1.18.0-rc.1
RecordExceptionis called on a span that is not recorded.(#7669)
See CHANGELOG for details.
NuGet: OpenTelemetry.Api.ProviderBuilderExtensions v1.18.0-rc.1
No notable changes.
See CHANGELOG for details.
NuGet: OpenTelemetry.Exporter.Console v1.18.0-rc.1
IEnumerable<KeyValuePair<string, object?>>). These attributes will be serialized as JSON objects.(#7015)
See CHANGELOG for details.
... (truncated)
1.18.0-beta.1
The following changes are from the previous release 1.17.0-beta.1.
NuGet: OpenTelemetry.Exporter.Prometheus.AspNetCore v1.18.0-beta.1
Fix concurrent scrapes returning an empty response under contention. Now the exporter will return an HTTP 500 error instead.
(#7571)
Waiting for concurrent scrapes to finish before collecting no longer blocks, which could stall concurrent scrapes being waited on.
(#7571)
Fixed the interaction between
PrometheusAspNetCoreOptions.TranslationStrategyand content negotiation. The configured strategy is now applied before content negotiation, instead of the negotiated escaping scheme replacing the strategy's, and theContent-Typeheader now reports the escaping scheme that was applied rather than the one that was negotiated.(#7610)
Fixed metric values and histogram bucket bounds being written with 17 significant digits instead of their shortest round-trippable representation.
(#7589)
Fixed the canonical representation used for histogram
leand summaryquantilelabel values falling back to 17 significant digits incorrectly.(#7589)
Fixed a race where a slow scrape could return an HTTP 200 instead of 408.
(#7615)
Updated OpenTelemetry core component version(s) to
1.18.0.(#7674)
See CHANGELOG for details.
NuGet: OpenTelemetry.Exporter.Prometheus.HttpListener v1.18.0-beta.1
Fix concurrent scrapes returning an empty response under contention. Now the exporter will return an HTTP 500 error instead.
(#7571)
Waiting for concurrent scrapes to finish before collecting no longer blocks, which could stall concurrent scrapes being waited on.
(#7571)
A scrape which is still collecting when the listener is disposed now returns an HTTP 503 response.
(#7587)
Shutting down the listener no longer waits indefinitely for its request processing loop to stop, and no longer throws if the loop faulted.
(#7587)
Fixed the interaction between
PrometheusHttpListenerOptions.TranslationStrategyand content negotiation. The configured strategy is now applied before content negotiation, instead of the negotiated escaping scheme replacing the strategy's, and theContent-Typeheader now reports the escaping scheme that was applied rather than the one that was negotiated.(#7610)
Fixed metric values and histogram bucket bounds being written with 17 significant digits instead of their shortest round-trippable representation.
(#7589)
Fixed the canonical representation used for histogram
leand summaryquantilelabel values falling back to 17 significant digits incorrectly.(#7589)
... (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 rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)