Allow loggers to be passed in #902
Merged
Conversation
This file contains 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
Motivation: The community recently discussed how logging should be done for the `AsyncHTTPClient` project. One of the outcomes was that frameworks should not log unless they are explicitly passed a logger. This allows applications to decide whether they care about logging for that component and have more control over the level at which that component logs at. Modifications: - Client allows for loggers in two places: one for background activity (connectivity state management and so on) and another for RPCs - Server allows a single logger to be specificed - By default all loggers are no-op loggers - Added 'none' option for 'CallOptions.RequestIDProvider' for cases where the request ID is already present on the logger Result: - gRPC does not log by default - Note: this includes tests, this will be addressed in a separate PR to keep PR size down
Lukasa
reviewed
Jul 20, 2020
| @@ -536,7 +549,8 @@ extension ClientConnection { | |||
| connectionKeepalive: ClientConnectionKeepalive = ClientConnectionKeepalive(), | |||
| connectionIdleTimeout: TimeAmount = .minutes(5), | |||
| callStartBehavior: CallStartBehavior = .waitsForConnectivity, | |||
| httpTargetWindowSize: Int = 65535 | |||
| httpTargetWindowSize: Int = 65535, | |||
| backgroundActivityLogger: Logger = Logger(label: "io.grpc", factory: { _ in SwiftLogNoOpLogHandler() }) | |||
There was a problem hiding this comment.
Is it worth encapsulating this in a static func on Logger we define internally, just so we have one-and-only-one place to write this code?
There was a problem hiding this comment.
It has to be public as it's used in a public signature. I wasn't sure making it underscored-internal was worth it in this case.
Sources/GRPC/ClientOptions.swift
Outdated
| public init( | ||
| customMetadata: HPACKHeaders = HPACKHeaders(), | ||
| timeLimit: TimeLimit = .none, | ||
| messageEncoding: ClientMessageEncoding = .disabled, | ||
| requestIDProvider: RequestIDProvider = .autogenerated, | ||
| requestIDProvider: RequestIDProvider = .none, |
There was a problem hiding this comment.
What's the reason for this change?
There was a problem hiding this comment.
It seemed more natural to opt-in to having request IDs generated rather than opting out. Not sure how much I agree with that thinking now. WDYT?
There was a problem hiding this comment.
I'm inclined to want to go to "fall into a pit of success": if users don't express a preference, giving them nice UUIDs makes their lives easier, and also our own during bug triage.
Sources/GRPC/PlatformSupport.swift
Outdated
| @@ -155,23 +153,22 @@ public enum PlatformSupport { | |||
| networkPreference: NetworkPreference = .best, | |||
| logger: Logger? = nil | |||
There was a problem hiding this comment.
Why are we defaulting the loggers in this file to nil as opposed to the no-op logger?
There was a problem hiding this comment.
I don't think there's a good reason to use an optional logger here; I just didn't think to change it.
glbrntt
added a commit
to glbrntt/grpc-swift
that referenced
this pull request
Jul 20, 2020
Motivation: In grpc#902 we changed behaviour such that gRPC will not log unless you explicitly pass in a logger. When tests fail it's useful to have the logs to triage them, especially when they fail under CI, for example. Modifications: - Add a capturing log handler to the test module which captures all emitted logs - Wire up loggers in tests - Hook up `GRPCTestCase` to print the captured logs only if a test fails (or an environment variable was set) - Add a bunch of missing `super.setUp` and `super.tearDown` calls Result: - If a test fails we get all `.trace` level logs emitted during that test
glbrntt
added a commit
to glbrntt/grpc-swift
that referenced
this pull request
Jul 20, 2020
Motivation: In grpc#902 we changed behaviour such that gRPC will not log unless you explicitly pass in a logger. When tests fail it's useful to have the logs to triage them, especially when they fail under CI, for example. Modifications: - Add a capturing log handler to the test module which captures all emitted logs - Wire up loggers in tests - Hook up `GRPCTestCase` to print the captured logs only if a test fails (or an environment variable was set) - Add a bunch of missing `super.setUp` and `super.tearDown` calls Result: - If a test fails we get all `.trace` level logs emitted during that test
glbrntt
added a commit
to glbrntt/grpc-swift
that referenced
this pull request
Jul 20, 2020
Motivation: In grpc#902 we changed behaviour such that gRPC will not log unless you explicitly pass in a logger. When tests fail it's useful to have the logs to triage them, especially when they fail under CI, for example. Modifications: - Add a capturing log handler factory to the test module which captures all logs emitted by handlers it makes - Wire up loggers in tests - Hook up `GRPCTestCase` to print the captured logs only if a test fails (or an environment variable was set) - Add a bunch of missing `super.setUp` and `super.tearDown` calls Result: - If a test fails we get all `.trace` level logs emitted during that test
Merged
glbrntt
added a commit
that referenced
this pull request
Jul 20, 2020
Motivation: In #902 we changed behaviour such that gRPC will not log unless you explicitly pass in a logger. When tests fail it's useful to have the logs to triage them, especially when they fail under CI, for example. Modifications: - Add a capturing log handler factory to the test module which captures all logs emitted by handlers it makes - Wire up loggers in tests - Hook up `GRPCTestCase` to print the captured logs only if a test fails (or an environment variable was set) - Add a bunch of missing `super.setUp` and `super.tearDown` calls Result: - If a test fails we get all `.trace` level logs emitted during that test
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Motivation:
The community recently discussed how logging should be done for the
AsyncHTTPClientproject. One of the outcomes was that frameworksshould not log unless they are explicitly passed a logger. This allows
applications to decide whether they care about logging for that
component and have more control over the level at which that component
logs at.
Modifications:
(connectivity state management and so on) and another for RPCs
where the request ID is already present on the logger
Result:
keep PR size down