feat(bungee): new skeleton for bungee plugin #84
Closed
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
eduv09
changed the title
|
@eduv09 nice job! You're getting to grips with this! Just one comment overall: The same though can be applied here for BUNGEE. Our idea is to implement a Strategy Pattern, such that the plugin defines the common functionality (e.g., generateLedgerStates) and each strategy (FabricStrategy, BesuStrategy, etc..) implements the specific function. Therefore, the BesuBungee and FabricBungee plugins would not exist anymore, along with the factory implementation. This should be fairly easy to change given that the code is already split in half, according to the ledger, but let us know if you need further input P.S. usually we give a bit more information on the commit message (and consequently the PR description) so that everyone knows what this commit is all about without having to figure out everything through the code. Take this commit as an example. And don't forget signing the commit using the |
AndreAugusto11
force-pushed
the
dev
branch
2 times, most recently
from
1fb18d2 to
50a07ba
Compare
1. We had dozens of different versions of ESLint, TSLint, Prettier declard in package.json files. Most versions were so old that they had ben EOL for several years too. 2. The presence of these and their scripts/config files/etc were making it impossible to run the linter on the sub-folders where these were being used. 3. As part of the big push to consolidate linting and code formatting on the entire project, this is phase 1 where I purged the mentioned dependencies from the project and now only the root package.json file declares them which is currently using the latest and greatest versions. 4. I know that the diff is huge in this pull request but there are no code changes that would cause bugs or logic modification at all. It's all just applying automated formatting and linting on sub-directories where previously this was impossible. E.g., this entire change should not cause any sort of bugs (but of course we'll see what the CI execution has to say about that). 5. The follow-up phases will consist of slowly applying more brazen refactoring in a much more targeted and surgical manner with small pull requests that are easy and quick to review. 6. The smaller pull requests will target sub-directories that we still cannot include in the linting (see the .eslintignore file for a list) and then culminate in us finally enabling the linter for everything. 7. The root folder's ESLint ignore file was refactored in a way that the files that we plan on fixing later are now ignored on a one by one basis instead of using patterns. This makes the ignore file a collection of to-dos where one can delete a single line from the ignore file (pertaining to a single file that is being ignored) and then proceed to fix the linter issues and send a small, easy to review pull request that does not overwhelm the maintainers when they are trying to review a huge diff with a large batch of linter fixes. Phase two of this project is then to one by one drain the ignore entries of these mentioned files from the ESLint ignore file. So, the bottom line is that this might look like a hugely disruptive change, but I've done my best to keep it as simple as possible so that we don't have to worry about reviewing it line by line and instead just focus on the big picture and the principle idea(s) behind it (code quality). If it does end up introducing a bug or CI breakage, I'll volunteer to fix those because I'm very confident that even if this does happen, it will be very minor and quick fixes necessary only so I don't mind being on the hook for it. Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
No changes other than the mass find & replacing of the GitHub actions that we use for setting up NodeJS, checking code out and caching. Address the warnings that started popping up in the CI that look like this, asking for a mass-upgrade of CI action versions: > Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: > actions/setup-node@v4.0.2, > actions/checkout@v4.1.1, > actions/cache@v4.0.1 > For more information see: > https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/. Fixes hyperledger-cacti#3079 Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
eduv09
force-pushed
the
Verifiable-private-proof-formats-SATP-1
branch
from
f04b30a to
456cb83
Compare
1. This does not upgrade all the container images that we build and publish as part of the project, it just upgrades the runner images that are being used by the continuous integration environment. 2. I also deleted a .yaml file that I left in the diff of another pull request by accident. Apologies for the littering. 3. Upgrade: as-in, I bumped the 20.04 versions to 22.04 4. Pin: as-in, replaced `ubuntu-latest` with `ubuntu-22.04` everywhere. 5. Reasoning: 24.04 is out now and so we have to keep up with the times because soon there'll be a stop to security patches even on what used to be LTS (20.04). It is better to find out now if we have a problem with ubuntu 22.04 rather than later when it's become urgent to upgrade. 6. The risk of something breaking because of these upgrades is most likely low since ubuntu is one of the most stable distributions out there. Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
|
This seems great work. Can you please add a README documenting the package with as much detail as possible? |
|
@RafaelAPB what do you think of also adding this to the CI? |
Bumps [crypto-js](https://github.com/brix/crypto-js) from 4.1.1 to 4.2.0. - [Commits](brix/crypto-js@4.1.1...4.2.0) --- updated-dependencies: - dependency-name: crypto-js dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com>
eduv09
force-pushed
the
Verifiable-private-proof-formats-SATP-1
branch
from
26b6be0 to
c73520b
Compare
eduv09
changed the title
1. This fixes many flakes that we were suffering from before with the older versions of the AIO image. 2. There could still be other flakes lurking around, but their numbers should definitely be going down with this change due to the increased stability of the new AIO image. Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
eduv09
force-pushed
the
Verifiable-private-proof-formats-SATP-1
branch
2 times, most recently
from
0fc466e to
8396797
Compare
1. The default/suggested NodeJS version we use and recommend is v18 LTS but there were a lot of places in the code that still declared v14 and v16 both of which are now EOL. 2. This was/is working fine but eventually we'll hit some compiler issue where the upgrade will be forced so it's most likely wiser to do this little quality of life/housekeeping change now when it's not urgent. Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
Bumps [follow-redirects](https://github.com/follow-redirects/follow-redirects) from 1.14.9 to 1.15.6. - [Release notes](https://github.com/follow-redirects/follow-redirects/releases) - [Commits](follow-redirects/follow-redirects@v1.14.9...v1.15.6) --- updated-dependencies: - dependency-name: follow-redirects dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com>
1. This is a pre-requisite for a bigger change that is adding gRPC endpoint support to the plugins. 2. We've had gRPC support in the API server for a long time, but the plugins are not yet able to register their own gRPC services as of now. 3. The reason we need the newer version is because some types are not exported by the older version that we'll be using to implement the gRPC support for plugins. Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
eduv09
force-pushed
the
Verifiable-private-proof-formats-SATP-1
branch
3 times, most recently
from
33039a2 to
f2cebdb
Compare
1. The solution here was to migrate the image from Debian to Ubuntu because it seems to not have the same vulnerabilities as the lastest stable Debian image does, so the change itself is to move to Ubuntu 24.04 LTS. 2. Also upgraded the Rust toolchain to the current latest which mandated a couple of small code changes that are also added in this commit. The original security report from Trivy that we've discoverd on the CI: ┌─────────────┬───────────────┬──────────┬───────────────────┐ │ Library │ Vulnerability │ Severity │ Installed Version │ ├─────────────┼───────────────┼──────────┼───────────────────┤ │ libgnutls30 │ CVE-2024-0553 │ HIGH │ 3.6.7-4+deb10u11 │ │ │ │ │ │ └─────────────┴───────────────┴──────────┴───────────────────┘ ... ┬──────────────────┬───────────────────────────────────────────┐ │ Fixed Version │ Title │ ┼──────────────────┼───────────────────────────────────────────┤ │ 3.6.7-4+deb10u12 │ gnutls: incomplete fix for CVE-2023-5981 │ │ │ https://avd.aquasec.com/nvd/cve-2024-0553 │ ┴──────────────────┴───────────────────────────────────────────┘ Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
1. User-defined Typescript type-guard function that asserts whether a value or
object is a `@grpc/grpc-js` {Partial<StatusObject>} or not.
The reason why it checks for {Partial} is because all of the properties of
the {StatusObject} are defined as optional for some reason, hence we cannot
assume anything about those being present or not by default.
2. Therefore this method will just check if the `code` property is set or not
and return `true` or `false` based on that.
3. The above is also the reason why the name of the function is slightly more
verbose than your average user-defined type-guard that could be named just
"isGrpcStatusObject()" but we wanted to make sure that more specific type-guards
can be added later that check for other optional properities or for the
presence of all of them together.
Link to the status builder within grpc-js:
https://github.com/grpc/grpc-node/blob/master/packages/grpc-js/src/status-builder.ts
Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
…on 2.0.0 OpenAPI * created new messages in openAPI definition
Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> feat(bungee): add skeleton for bungee Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt> refactor(plugin-satp-hermes): update messages formats SATP core version 2.0.0 OpenAPI * created new messages in openAPI definition feat(bungee): fix bungee dependencies Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt> feat(SATP-Hermes): gateway coordinator feat(SATP-Hermes): gateway coordinator feat(SATP-Hermes): gateway coordinator WIP Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> test(SATP-Hermes): enable decorators needed for gateway coordinator checks Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> docs(SATP-Hermes): update package.json Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> refactor(SATP-Hermes): re-estructure project Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> refactor(SATP-Hermes): re-estructure project Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> refactor(SATP-Hermes): re-estructure project Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> refactor(SATP-Hermes): remove kotlin sdk, add protobuffer Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> feat(SATP-Hermes): protobuf structure Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> feat(SATP-Hermes): protobuf structure Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> refactor(SATP-Hermes): refactor naming Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> feat(SATP-Hermes): protobuf structure improvements Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt> feat(SATP-Hermes): update yarn.lock and clean up Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt> refactor(test-tooling): fix types of streams: use NodeJS.ReadableStream 1. The container management library that we use in the test infrastructure (called dockerode) is expecting streams that are defined in the global namespace of the `@types/node` library, e.g. the standard library of NodeJS itself. 2. Previously we were using the "streams" package to provide type information to the streams that we were passing around to dockerode and it was working fine, but after some changes that seem unrelated this has broken the compilation process. 3. The mentioned changes are not yet on the main branch, but we expect them to be there soon and so this change is laying the groundwork for that by pre-emptively fixing the broken build's root cause which is that the test-tooling package does not declare it's typings related dependencies correctly: It implicitly uses the NodeJS standard library's types but so far had not declared them on the package level. 4. This change is therefore to rectify the issue of the `@types/node` dependency missing from the test-tooling package and also the refactoring of some of the test ledger classes which were relying on the `streams` builtin package instead of correctly using the NodeJS.ReadableStream global. 5. Earlier the reasoning for this was that we try to avoid pulling in types from the global scope because we try to avoid any sort of dependency on the global scope in general. Once we have proof though that this is causing issues with the build, then we must give up the principle for practical reasons (and only in the minimum viable scope, e.g. this does not change the fact that everywhere else in the codebase we should still do our best to avoid using the global scoped classes, types, functions, etc..). Thank you to @AndreAugusto11 and @RafaelAPB for pointing out this issue through the pull request of his that is currently being worked on at the time of this writing: RafaelAPB#72 Related to but does not address hyperledger-cacti#2811 Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com> refactor(SATP-Hermes): remove unused packages * updated openapi version * updated bungee plugin version to 2.0.0-alpha.2 Co-authored-by: Peter Somogyvari <peter.somogyvari@accenture.com> Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt> Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com>
Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
- Add SATP-Hermes to CI - Fix some SATP tests that were hanging the terminal after execution - Migrated SATP tests to Fabric 2.5.x LTS image The older Fabric AIO images have become flaky unfortunately due to auto-upgrades that we can't pin down due to upstream repos doing it. Co-authored-by: Peter Somogyvari <peter.somogyvari@accenture.com> Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com> Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt>
Signed-off-by: André Augusto <andre.augusto@tecnico.ulisboa.pt>
Signed-off-by: Rafael Belchior <rafael.belchior@tecnico.ulisboa.pt>
* plugin follows strategy design pattern defined in folder strategy/ * current version has strategy for fabric and besu networks * includes a few tests to demonstrate basic functionality * added README with package description Signed-off-by: eduv09 <eduardovasques10@tecnico.ulisboa.pt>
eduv09
force-pushed
the
Verifiable-private-proof-formats-SATP-1
branch
from
f2cebdb to
7cf0c90
Compare
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.
New architecture for the BUNGEE Plugin for Fabric and Besu networks.