Add Distributed Tracing support #177

slashmo · 2025-08-05T21:02:52Z

Motivation

To make the most of Distributed Tracing, users may want to view operations performed by their database clients as part of a larger request to gain more insight into their distributed systems.

Modifications

Add new DistributedTracingSupport trait, enabled by default
Create client span in single and pipelined command execution
Updated the command builder to generate a command name property, used in the db.operation.name span attribute
Add open-telemetry example to showcase Distributed Tracing support

When implementing the span and span attributes I closely followed the OpenTelemetry semantic conventions for Redis and general database client spans:

I believe that this provides the highest chance of the span being recognized as a database client operation by various observability backends.

Result

Users can now use Distributed Tracing spans automatically created by the library to more deeply observe their systems.

Closes #176

czechboy0

Excited for distributed tracing 👏

czechboy0 · 2025-08-06T06:37:07Z

Sources/Valkey/Connection/ValkeyConnection.swift

@@ -161,16 +176,48 @@ public final actor ValkeyConnection: ValkeyClientProtocol, Sendable {

    @inlinable
    func _execute<Command: ValkeyCommand>(command: Command) async throws -> RESPToken {
+        #if DistributedTracingSupport
+        let span = startSpan(Command.name, ofKind: .client)
+        defer { span.end() }


If you wanted to get rid of the unstructured span, could also do something like:

func _execute<Command: ValkeyCommand>(command: Command) async throws -> RESPToken { #if DistributedTracingSupport try await _executeWithTracing(command: command) #else try await _execute0(command: command) #endif } func _execute0<Command: ValkeyCommand>(command: Command) async throws -> RESPToken { // existing execute logic } #if DistributedTracingSupport func _executeWithTracing<Command: ValkeyCommand>(command: Command) async throws -> RESPToken { try await withSpan { span in // update the span do { try await _execute0(command: command) } catch { // extract info from the error into the span } } #endif

Thanks for the suggestion. This is similar to what I originally went with. I eventually switched to the more imperative API because I thought it was easier to read. I'm happy to explore withSpan again though.

I'll let Adam/Fabian steer you here, I'm also happy with either.

Pretty sure the client fast path API will not use structured concurrency. See #158.

codecov-commenter · 2025-08-06T07:54:19Z

Codecov Report

❌ Patch coverage is 96.10390% with 3 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (main@cab1563). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
Sources/Valkey/Connection/ValkeyConnection.swift	96.61%	2 Missing ⚠️
Sources/Valkey/ValkeyConnectionFactory.swift	83.33%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #177   +/-   ##
=======================================
  Coverage        ?   39.90%           
=======================================
  Files           ?       86           
  Lines           ?    15277           
  Branches        ?        0           
=======================================
  Hits            ?     6097           
  Misses          ?     9180           
  Partials        ?        0

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

slashmo · 2025-08-06T08:27:41Z

The API breakage check rightfully complains about the added protocol requirement ValkeyCommand.name. Is this acceptable given we're pre-1.0 or would you like me to add a default in an extension, for example like this:

extension ValkeyCommand {
    public static var name: String { String(describing: Self.self) }
}

adam-fowler · 2025-08-06T09:01:49Z

The API breakage check rightfully complains about the added protocol requirement ValkeyCommand.name. Is this acceptable given we're pre-1.0 or would you like me to add a default in an extension, for example like this:
extension ValkeyCommand {

    public static var name: String { String(describing: Self.self) }

}

API breakages are fine

adam-fowler

A couple of initial comments.

It's a shame I haven't got the benchmarks running for non-members of the repo. It fails on add the comment to the PR. I imagine this will add allocations. Can you run the benchmarks locally (run main and compare this PR with main) and post the results as a comment, cheers.

Sources/Valkey/Connection/ValkeyConnection.swift

Tests/ValkeyTests/Utils/TestTracer.swift

Examples/open-telemetry/Package.swift

slashmo · 2025-08-06T15:38:42Z

Here's the benchmark diff between main and the pr. I ran them with Swift 6.2, let me know if you'd prefer 6.1 and I can install it:

swift-driver version: 1.127.10 Apple Swift version 6.2 (swiftlang-6.2.0.14.8 clang-1700.3.14.6)
Target: arm64-apple-macosx15.0

===============================================================
Threshold deviations for ValkeyBenchmarks:Client: GET benchmark
===============================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Malloc (total) (K, %)                    │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              74 │              81 │              10 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │              77 │              83 │               8 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │              78 │              85 │               8 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

=========================================================================================================
Threshold deviations for ValkeyBenchmarks:Client: GET benchmark | parallel 20 | 20 concurrent connections
=========================================================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Malloc (total) (K, %)                    │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p75                                      │             117 │             125 │               6 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

=========================================================================================================
Threshold deviations for ValkeyBenchmarks:Client: GET benchmark | parallel 50 | 20 concurrent connections
=========================================================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Malloc (total) (K, %)                    │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │               5 │               7 │              30 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │               8 │              11 │              45 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │              11 │              18 │              59 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Instructions (M, %)                      │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │             247 │             269 │               8 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │             249 │             271 │               9 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │             250 │             272 │               9 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

===================================================================
Threshold deviations for ValkeyBenchmarks:Connection: GET benchmark
===================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Throughput (# / s) (#, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              29 │              27 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │              29 │              27 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │              29 │              27 │               6 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Time (wall clock) (ms, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              34 │              37 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │              35 │              37 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │              35 │              38 │               8 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Time (total CPU) (ms, %)                 │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              29 │              31 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │              29 │              31 │               6 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │              30 │              32 │               9 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Instructions (M, %)                      │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │             225 │             247 │               9 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │             225 │             247 │               9 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p75                                      │             225 │             247 │               9 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

========================================================================
Threshold deviations for ValkeyBenchmarks:Connection: Pipeline benchmark
========================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Throughput (# / s) (#, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              17 │              16 │               5 │               5 │
├──────────────────────────────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤
│ p50                                      │              17 │              16 │               5 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Malloc (total) (K, %)                    │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p25                                      │              14 │              26 │              83 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

====================================================================
Threshold deviations for ValkeyBenchmarks:HashSlot – {user}.whatever
====================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Throughput (# / s) (#, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p75                                      │              57 │              53 │               7 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Time (wall clock) (ms, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p75                                      │              18 │              19 │               7 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

=====================================================================================
Threshold deviations for ValkeyBenchmarks:ValkeyCommandEncoder – Command with 7 words
=====================================================================================
╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Time (wall clock) (μs, %)                │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p75                                      │             420 │             441 │               5 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

╒══════════════════════════════════════════╤═════════════════╤═════════════════╤═════════════════╤═════════════════╕
│ Time (total CPU) (μs, %)                 │            main │              pr │    Difference % │     Threshold % │
╞══════════════════════════════════════════╪═════════════════╪═════════════════╪═════════════════╪═════════════════╡
│ p75                                      │             421 │             443 │               5 │               5 │
╘══════════════════════════════════════════╧═════════════════╧═════════════════╧═════════════════╧═════════════════╛

New baseline 'pr' is WORSE than the 'main' baseline thresholds.

error: benchmarkThresholdRegression

fabianfett

@slashmo Thank you for taking a stab at this one! Really like the direction. We will need to dive deep on some of the aspects.

fabianfett · 2025-08-20T11:55:54Z

Sources/Valkey/Connection/ValkeyConnection.swift

+                while prefixEndIndex < message.endIndex, message[prefixEndIndex] != " " {
+                    message.formIndex(after: &prefixEndIndex)
+                }
+                let prefix = message[message.startIndex..<prefixEndIndex]


Pretty sure all of this is in the UTF8 space. Can we operate on the utf8 view instead?

Maybe break the code into a separate function as well, to make it clear what you are doing

fabianfett · 2025-08-20T11:57:10Z

Sources/Valkey/Connection/ValkeyConnection.swift

+        attributes["network.peer.address"] = channel.remoteAddress?.ipAddress
+        attributes["network.peer.port"] = channel.remoteAddress?.port
+        attributes["server.address"] = address?.hostOrSocketPath
+        attributes["server.port"] = address?.port == 6379 ? nil : address?.port


I think we should check first if we have a value. Also please use explicit self. Can we make all those labels configurable?

These look like the OTel Semantic Conventions, this could also take an explicit dependency on that library to make these typesafe.

I think we should check first if we have a value.

Functionally there's no difference between setting a nil value for an attribute and not setting it at all. Or do you see a performance optimization in not going through the subscript unless necessary?

Tests/ValkeyTests/Utils/TestTracer.swift

Sources/Valkey/Connection/ValkeyConnection.swift

slashmo · 2025-08-21T17:19:04Z

After discussing this PR in more detail, @fabianfett and I defined the following next actions:

Split off command name generating into separate PR (Add static name to Valkey commands #186)
Move TestTracer into swift-distributed-tracing
Pass (any Tracer)? to ValkeyClient and ValkeyClusterClient instead of using global
Keep the initial PR focused on just creating a span in ValkeyConnection (Future PRs may wrap this span in other spans created by ValkeyClient and ValkeyClusterClient to include operations such as node discovery and establishing a connection)
Configurable span attribute names and span operation name
Make as few allocations as possible and generally create as little overhead as possible when the DistributedTracingSupport trait is enabled

adam-fowler · 2025-08-21T20:46:34Z

This list looks fairly good. You did forget find a solution for pipelined commands.

slashmo · 2025-08-21T20:56:21Z

You did forget find a solution for pipelined commands.

We also briefly discussed this, forgot to mention it here. What I'd ideally like to do is to pass whether the pipelined execute was preceded by a MULTI to the variadic generics overload of execute so that it can either create a span called MULTI or a span called something like pipeline with a span attribute. Valkey Swift's transactional API would pass true to the execute method but by default it'd be false. In either case, the span created in execute should have a span attribute with the list of pipelined command names.

ktoso · 2025-08-27T08:11:12Z

Great to see this happening, thank you @slashmo, @adam-fowler, @fabianfett! Was just discussing otel plans with Moritz and the thread came up.

I'm really glad to see this approach does:

depend on swift-distributed-tracing directly, that's the way we should be doing this
guards the feature with a package trait
is on by default; I think that's the right choice, otherwise running around and trying to enable all the possible traits would be tricky.

It's somewhat concerning to adopt a pattern of "pass (any Tracer)?" because that's somewhat incompatible with any of the ways to use tracer.withSpan { ... }. You can't really:

// bad, always need to run the nested code
tracer?.withSpan(...) { span in code } // ofc wrong, always need to run the 'code'

// so what... duplication like this is absolutely not ok
if tracer {  
  tracer.withSpan(...) { span in code } 
} else {
  code 
}

While with noop-tracers this "just works" because:

tracer.withSpan(...) { noopSpan in code } // noopspan is a struct

Are we absolutely certain this optional-nil vs "noop" is a must here? Is the difference actually significant enough to warrant the pain, given this is in tracing is enabled mode? If so, we must answer the question what libraries should be doing when using tracer like this.

Because what would the pattern look like for adopter code of tracing that do use structured concurrency. I understand your specific lib is not using structured concurrency in the core, but if we start recommending "pass (any Tracer)?" I'm concerned this will become a pattern everywhere, and just not work at all for libs doing structured concurrency

We have a weekly call we could purpose for designing the answer here, @slashmo will try to organize the invites! :-)

ktoso · 2025-08-27T12:10:43Z

Seems we had a minor misunderstanding here!

Coming off with a chat with @fabianfett - we clarified this isn't trying to establish "the" pattern we'll ask other libraries to adopt. So technically it's really up to this lib to do whatever you want with passing the tracer -- optional or noop. So my review isn't as strict really, we were only concerned about end users adopting this pattern in their own code, but this isn't what this valkey intended to be a showcase of.

The optional vs no-op difference is rather minimal in performance but it does cause a few rare allocations whereas without it valkey is solidly at 0 by the way;

The way this sets up the tracer "once" in the config object is also great and I have no concerns there.

We should bring the "in memory recording" tracer to swift-distributed-tracing as discussed up-thread though, I look forward to accepting that over there :-)

In case it didn't get across the first comment: this is GREAT and I'm really happy with this PR, I just had some usability concerns if we were to recommend this style to everyone (but we don't, so there's no problem) 👍

ktoso

Looks good! I think that's a nice approach, maybe just add some comments about it

ktoso · 2025-08-27T13:05:08Z

Sources/Valkey/Connection/ValkeyConnectionConfiguration.swift

@@ -117,6 +120,10 @@ public struct ValkeyConnectionConfiguration: Sendable {
    /// Default value is `nil` (no client name is set).
    public var clientName: String?

+    #if DistributedTracingSupport
+    public var tracing: ValkeyTracingConfiguration = .init()


On this one maybe just a small note that picks global bootstrapped tracer at this point in time and allows configuring tracing?

ktoso · 2025-08-27T13:05:10Z

Sources/Valkey/Connection/ValkeyConnectionConfiguration.swift

+#if DistributedTracingSupport
+public struct ValkeyTracingConfiguration: Sendable {
+
+    public var tracer: (any Tracer)? = InstrumentationSystem.tracer


Probably worth some docs here how this is

by default picks up global tracer bootstrapped in InstrumentationSystem

can be set to nil to disable tracing entirely, even if global tracer is bootstrapped

can be set to specific tracer, which takes precedence over global bootstrapped tracer

?

If global tracer is NoOp tracer is it worth setting tracer to nil?

I had asked about this over here #177 (comment) and had a chat with @fabianfett about it.

So there seems to be minor differences and although the performance is roughly the same it seems to have caused an odd allocation here and there, whereas without it the allocation count in valkey was a solid 0... The NoopTracer is definitely nicer to use but if you truly want to go with the Tracer? here I think that's fine -- @fabianfett strongly preferred that at least for this lib.

If we were to suggest people in libraries, I think you're very right -- doing Tracer = NoopTracer... is preferable, rather than Tracer? because of the usability aspect, but in this specific lib I think it's fine... Since Fabian really wanted to have the allocs count a strict zero here heh :)

It's not that the allocations come from the noop-spans btw, they are structs... so I'm uncertain where they were coming from; we'd need to do more digging

@slashmo Can we get some comments on ValkeyTracingConfiguration and its members. This is currently a public symbol without any comments.

Yes, let me add some.

adam-fowler

In general I think we are good, added a few more comments.

Sources/Valkey/Connection/ValkeyConnection.swift

adam-fowler · 2025-08-28T17:03:43Z

Sources/Valkey/Connection/ValkeyConnection.swift

-                self.channelHandler.write(command: command, continuation: continuation, requestID: requestID)
+        } catch let error as ValkeyClientError {
+            #if DistributedTracingSupport
+            span?.recordError(error)


Maybe wrap all this code in if let span

adam-fowler · 2025-08-28T17:06:11Z

Sources/Valkey/Connection/ValkeyConnectionConfiguration.swift

+#if DistributedTracingSupport
+public struct ValkeyTracingConfiguration: Sendable {
+
+    public var tracer: (any Tracer)? = InstrumentationSystem.tracer


If global tracer is NoOp tracer is it worth setting tracer to nil?

adam-fowler · 2025-08-28T17:06:47Z

Sources/Valkey/Connection/ValkeyConnection.swift

+                while prefixEndIndex < message.endIndex, message[prefixEndIndex] != " " {
+                    message.formIndex(after: &prefixEndIndex)
+                }
+                let prefix = message[message.startIndex..<prefixEndIndex]


Maybe break the code into a separate function as well, to make it clear what you are doing

Package.swift

adam-fowler

You need to ensure everything runs with the DistributedTracingSupport trait disabled

swift test --disable-default-traits

Sources/Valkey/Connection/ValkeyConnection.swift

Benchmarks/ValkeyBenchmarks/ValkeyConnectionBenchmark.swift

Signed-off-by: Moritz Lang <[email protected]>

Signed-off-by: Fabian Fett <[email protected]>

Signed-off-by: Moritz Lang <[email protected]>

adam-fowler

One minor thing and we're good to go

adam-fowler · 2025-09-13T15:27:17Z

Sources/Valkey/Connection/ValkeyConnectionConfiguration.swift

+#if DistributedTracingSupport
+public struct ValkeyTracingConfiguration: Sendable {
+
+    public var tracer: (any Tracer)? = InstrumentationSystem.tracer


@slashmo Can we get some comments on ValkeyTracingConfiguration and its members. This is currently a public symbol without any comments.

Signed-off-by: Moritz Lang <[email protected]>

Sources/Valkey/Connection/ValkeyConnection.swift

Signed-off-by: Moritz Lang <[email protected]>

adam-fowler

Thanks @slashmo this is great

slashmo force-pushed the feature/distributed-tracing branch from 6c6fb74 to 5606d42 Compare August 5, 2025 21:08

czechboy0 reviewed Aug 6, 2025

View reviewed changes

slashmo force-pushed the feature/distributed-tracing branch from a67131f to c57a3b5 Compare August 6, 2025 08:29

adam-fowler reviewed Aug 6, 2025

View reviewed changes

Sources/Valkey/Connection/ValkeyConnection.swift Outdated Show resolved Hide resolved

Sources/Valkey/Connection/ValkeyConnection.swift Outdated Show resolved Hide resolved

adam-fowler reviewed Aug 6, 2025

View reviewed changes

Tests/ValkeyTests/Utils/TestTracer.swift Outdated Show resolved Hide resolved

Examples/open-telemetry/Package.swift Outdated Show resolved Hide resolved

fabianfett reviewed Aug 20, 2025

View reviewed changes

slashmo mentioned this pull request Aug 21, 2025

Add static name to Valkey commands #186

Merged

fabianfett added this to the 1.0 milestone Aug 25, 2025

fabianfett force-pushed the feature/distributed-tracing branch from 9129908 to 3d01310 Compare August 27, 2025 12:56

ktoso approved these changes Aug 27, 2025

View reviewed changes

fabianfett force-pushed the feature/distributed-tracing branch from 8030400 to 56ed070 Compare August 28, 2025 10:13

StefanPrintezis mentioned this pull request Aug 28, 2025

Investigate adopting Tracing vapor/postgres-nio#402

Open

adam-fowler reviewed Aug 28, 2025

View reviewed changes

adam-fowler reviewed Sep 4, 2025

View reviewed changes

Package.swift Outdated Show resolved Hide resolved

adam-fowler reviewed Sep 9, 2025

View reviewed changes

Sources/Valkey/Connection/ValkeyConnection.swift Show resolved Hide resolved

Sources/Valkey/Connection/ValkeyConnection.swift Show resolved Hide resolved

Benchmarks/ValkeyBenchmarks/ValkeyConnectionBenchmark.swift Show resolved Hide resolved

slashmo added 6 commits September 13, 2025 16:07

Add static name to Valkey commands

14d5172

Signed-off-by: Moritz Lang <[email protected]>

Add Distributed Tracing support behind new trait

8560eca

Signed-off-by: Moritz Lang <[email protected]>

Add open-telemetry example showcasing Distributed Tracing support

061f031

Signed-off-by: Moritz Lang <[email protected]>

Run swift-format

c95c599

Signed-off-by: Moritz Lang <[email protected]>

Document open-telemetry example

ef47244

Signed-off-by: Moritz Lang <[email protected]>

Remove left-over TODO

a4e8772

Signed-off-by: Moritz Lang <[email protected]>

slashmo and others added 15 commits September 13, 2025 16:07

Re-format docker-compose.yml

db86f71

Signed-off-by: Moritz Lang <[email protected]>

Re-format ValkeyConnectionTests.swift

a6e27b7

Signed-off-by: Moritz Lang <[email protected]>

Clean up service bootstrap in open-telemetry example

29d5ba0

Signed-off-by: Moritz Lang <[email protected]>

Remove open-telemetry example

a63dc6b

Signed-off-by: Moritz Lang <[email protected]>

Improve tracing support

baff60c

Signed-off-by: Fabian Fett <[email protected]>

Fix tests

d1b16e7

Signed-off-by: Fabian Fett <[email protected]>

swift-format

ada0aa2

Signed-off-by: Fabian Fett <[email protected]>

swift-format

784f53e

Signed-off-by: Fabian Fett <[email protected]>

Use TestTracer from TracingTestKit

5b59b04

Signed-off-by: Moritz Lang <[email protected]>

Put ValkeyConnection.address behind Tracing trait

4d0066e

Signed-off-by: Moritz Lang <[email protected]>

Extract simple error prefix into computed variable

f1667d6

Signed-off-by: Moritz Lang <[email protected]>

Add missing availability macro to ValkeyTracingConfiguration

a1b605f

Signed-off-by: Moritz Lang <[email protected]>

Switch to swift-distributed-tracing upstream

960c035

Signed-off-by: Moritz Lang <[email protected]>

Fix compilation without DistributedTracingSupport trait

bb8839b

Signed-off-by: Moritz Lang <[email protected]>

Fix benchmark compilation

34db440

Signed-off-by: Moritz Lang <[email protected]>

slashmo force-pushed the feature/distributed-tracing branch from 6ca1b36 to 34db440 Compare September 13, 2025 14:27

adam-fowler reviewed Sep 13, 2025

View reviewed changes

Document tracing configuration

faf9dcc

Signed-off-by: Moritz Lang <[email protected]>

slashmo force-pushed the feature/distributed-tracing branch from 9643e52 to faf9dcc Compare September 13, 2025 17:40

czechboy0 reviewed Sep 13, 2025

View reviewed changes

Sources/Valkey/Connection/ValkeyConnection.swift Outdated Show resolved Hide resolved

Set span status code for all errors

60c0a01

Signed-off-by: Moritz Lang <[email protected]>

adam-fowler approved these changes Sep 14, 2025

View reviewed changes

adam-fowler merged commit 064e92f into valkey-io:main Sep 14, 2025
12 of 13 checks passed

slashmo deleted the feature/distributed-tracing branch September 14, 2025 12:01

Add Distributed Tracing support #177

Add Distributed Tracing support #177

Conversation

slashmo commented Aug 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Modifications

Result

Uh oh!

czechboy0 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codecov-commenter commented Aug 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

slashmo commented Aug 6, 2025

Uh oh!

adam-fowler commented Aug 6, 2025

Uh oh!

adam-fowler left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

slashmo commented Aug 6, 2025

Uh oh!

fabianfett left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

slashmo commented Aug 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adam-fowler commented Aug 21, 2025

Uh oh!

slashmo commented Aug 21, 2025

Uh oh!

ktoso commented Aug 27, 2025

Uh oh!

ktoso commented Aug 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ktoso left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

adam-fowler left a comment

slashmo commented Aug 5, 2025 •

edited

Loading

codecov-commenter commented Aug 6, 2025 •

edited

Loading

slashmo commented Aug 21, 2025 •

edited

Loading

ktoso commented Aug 27, 2025 •

edited

Loading