Implement tracking of known resource names by caller stream. Fixes #399 (Attempt #2) #508
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
|
@alecholmez Please check. |
|
Initial PR and review comments are in #490 |
|
@kkalin68 can you do me a favor and rebase/squash these commits? Just approved the build to run |
…voyproxy#399 Signed-off-by: Konstantin Kalin <[email protected]>
|
@alecholmez Done. |
alecholmez
approved these changes
Oct 25, 2021
There was a problem hiding this comment.
I think this is good to go know @kkalin68. Thanks for the PR and working with me in the review process
valerian-roche
added a commit
to valerian-roche/go-control-plane
that referenced
this pull request
Jan 19, 2024
…e stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This PR replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to valerian-roche/go-control-plane
that referenced
this pull request
Jan 19, 2024
…e stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This PR replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to valerian-roche/go-control-plane
that referenced
this pull request
Jan 22, 2024
…e stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This PR replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to DataDog/go-control-plane
that referenced
this pull request
Feb 5, 2024
…e stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This PR replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to DataDog/go-control-plane
that referenced
this pull request
Feb 16, 2024
…e stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This PR replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to DataDog/go-control-plane
that referenced
this pull request
Feb 16, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to DataDog/go-control-plane
that referenced
this pull request
Feb 23, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
zhiyanfoo
pushed a commit
to DataDog/go-control-plane
that referenced
this pull request
Apr 10, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to DataDog/go-control-plane
that referenced
this pull request
Apr 10, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to valerian-roche/go-control-plane
that referenced
this pull request
May 9, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
valerian-roche
added a commit
to valerian-roche/go-control-plane
that referenced
this pull request
Aug 8, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
shamdor
pushed a commit
to DataDog/go-control-plane
that referenced
this pull request
Sep 23, 2024
…delta (#9) Currently sotw and delta watches use different interfaces to track the stream state An assumption used to be that sotw watches are stateless and can be processed from the request itself. This was already altered in envoyproxy#508, but it only covers the usecases partially (e.g. it does not cover subscribing/unsubscribing to wildcard). The current object passed to track the subscription provides multiple issues: - it is not an interface, restricting evolution or testing for users building a separate cache - it tracks some objects per types, other globally, based on sotw or delta, making it confusing and hard to use in a consistent way This commit replaces the legacy streamState by the notion of Subscription per type. StreamState is not relevant to the cache, as it only requires to know the subscription status (e.g. the last nonce is not needed), while a stream can multiplex multiple types. Further work will come to: - fill the subscription with the same attributes in sotw and delta, allowing to take proper decisions on what to reply - Fix multiple issues within the cache in sotw - Fix handling of explicit vs. legacy wildcard in sotw and delta Now that the `Subscription` has been standardized to be uniform for `sotw` and `delta`, as well as always for the given request type, this PR ensures all the fields are properly populated in all cases A follow-up PR will rely on those informations to address current issues in the cache implementations related to resource tracking Signed-off-by: Valerian Roche <[email protected]>
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.
No description provided.