cerc-io/lighthouse
15
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity
40c6daa34b
lighthouse/validator_client/slashing_protection/Cargo.toml

31 lines
760 B
TOML
Raw Normal View History

Implement Slashing Protection (#1116) * Implement slashing protection Roll-up of #588 with some conflicts resolved * WIP improvements * Require slot uniqueness for blocks (rather than epochs) * Native DB support for Slot and Epoch * Simplify surrounding/surrounded-by queries * Implement unified slashing protection database A single SQL database saves on open file descriptors. * Make slashing protection concurrency safe. Revive tests, add parallel tests. * Some simplifications * Auto-registration, test clean-ups * More tests, clean-ups, hardening * Fix comments in BLS * Optimise bulk validator registration * Delete outdated tests * Use bundled SQLite in slashing protection * Auto-register validators in simulation * Use real signing_root in slashing protection * Update book for --auto-register * Refine log messages and help flags * Correct typo in Cargo.toml authors * Fix merge conflicts * Safer error handling in sqlite slot/epoch * Address review comments * Add attestation test mutating block root Co-authored-by: pscott <scottpiriou@gmail.com>
2020-05-18 06:25:16 +00:00
[package]
name = "slashing_protection"
version = "0.1.0"
authors = ["Michael Sproul <michael@sigmaprime.io>", "pscott <scottpiriou@gmail.com>"]
Update to Rust 1.59 and 2021 edition (#3038) ## Proposed Changes Lots of lint updates related to `flat_map`, `unwrap_or_else` and string patterns. I did a little more creative refactoring in the op pool, but otherwise followed Clippy's suggestions. ## Additional Info We need this PR to unblock CI.
2022-02-25 00:10:17 +00:00
edition = "2021"
Implement standard keystore API (#2736) ## Issue Addressed Implements the standard key manager API from https://ethereum.github.io/keymanager-APIs/, formerly https://github.com/ethereum/beacon-APIs/pull/151 Related to https://github.com/sigp/lighthouse/issues/2557 ## Proposed Changes - [x] Add all of the new endpoints from the standard API: GET, POST and DELETE. - [x] Add a `validators.enabled` column to the slashing protection database to support atomic disable + export. - [x] Add tests for all the common sequential accesses of the API - [x] Add tests for interactions with remote signer validators - [x] Add end-to-end tests for migration of validators from one VC to another - [x] Implement the authentication scheme from the standard (token bearer auth) ## Additional Info The `enabled` column in the validators SQL database is necessary to prevent a race condition when exporting slashing protection data. Without the slashing protection database having a way of knowing that a key has been disabled, a concurrent request to sign a message could insert a new record into the database. The `delete_concurrent_with_signing` test exercises this code path, and was indeed failing before the `enabled` column was added. The validator client authentication has been modified from basic auth to bearer auth, with basic auth preserved for backwards compatibility.
2022-01-30 23:22:04 +00:00
autotests = false
[[test]]
name = "slashing_protection_tests"
path = "tests/main.rs"
Implement Slashing Protection (#1116) * Implement slashing protection Roll-up of #588 with some conflicts resolved * WIP improvements * Require slot uniqueness for blocks (rather than epochs) * Native DB support for Slot and Epoch * Simplify surrounding/surrounded-by queries * Implement unified slashing protection database A single SQL database saves on open file descriptors. * Make slashing protection concurrency safe. Revive tests, add parallel tests. * Some simplifications * Auto-registration, test clean-ups * More tests, clean-ups, hardening * Fix comments in BLS * Optimise bulk validator registration * Delete outdated tests * Use bundled SQLite in slashing protection * Auto-register validators in simulation * Use real signing_root in slashing protection * Update book for --auto-register * Refine log messages and help flags * Correct typo in Cargo.toml authors * Fix merge conflicts * Safer error handling in sqlite slot/epoch * Address review comments * Add attestation test mutating block root Co-authored-by: pscott <scottpiriou@gmail.com>
2020-05-18 06:25:16 +00:00
[dependencies]
tempfile = "3.1.0"
Directory Restructure (#1163) * Move tests -> testing * Directory restructure * Update Cargo.toml during restructure * Update Makefile during restructure * Fix arbitrary path
2020-05-18 11:24:23 +00:00
types = { path = "../../consensus/types" }
Upgrade dependencies (#2513) ## Proposed Changes * Consolidate Tokio versions: everything now uses the latest v1.10.0, no more `tokio-compat`. * Many semver-compatible changes via `cargo update`. Notably this upgrades from the yanked v0.8.0 version of crossbeam-deque which is present in v1.5.0-rc.0 * Many semver incompatible upgrades via `cargo upgrades` and `cargo upgrade --workspace pkg_name`. Notable ommissions: - Prometheus, to be handled separately: https://github.com/sigp/lighthouse/issues/2485 - `rand`, `rand_xorshift`: the libsecp256k1 package requires 0.7.x, so we'll stick with that for now - `ethereum-types` is pinned at 0.11.0 because that's what `web3` is using and it seems nice to have just a single version ## Additional Info We still have two versions of `libp2p-core` due to `discv5` depending on the v0.29.0 release rather than `master`. AFAIK it should be OK to release in this state (cc @AgeManning )
2021-08-17 01:00:24 +00:00
rusqlite = { version = "0.25.3", features = ["bundled"] }
Update external deps (#1711) ## Issue Addressed - Resolves #1706 ## Proposed Changes Updates dependencies across the workspace. Any crate that was not able to be brought to the latest version is listed in #1712. ## Additional Info NA
2020-10-05 08:22:19 +00:00
r2d2 = "0.8.9"
Upgrade dependencies (#2513) ## Proposed Changes * Consolidate Tokio versions: everything now uses the latest v1.10.0, no more `tokio-compat`. * Many semver-compatible changes via `cargo update`. Notably this upgrades from the yanked v0.8.0 version of crossbeam-deque which is present in v1.5.0-rc.0 * Many semver incompatible upgrades via `cargo upgrades` and `cargo upgrade --workspace pkg_name`. Notable ommissions: - Prometheus, to be handled separately: https://github.com/sigp/lighthouse/issues/2485 - `rand`, `rand_xorshift`: the libsecp256k1 package requires 0.7.x, so we'll stick with that for now - `ethereum-types` is pinned at 0.11.0 because that's what `web3` is using and it seems nice to have just a single version ## Additional Info We still have two versions of `libp2p-core` due to `discv5` depending on the v0.29.0 release rather than `master`. AFAIK it should be OK to release in this state (cc @AgeManning )
2021-08-17 01:00:24 +00:00
r2d2_sqlite = "0.18.0"
Update external deps (#1711) ## Issue Addressed - Resolves #1706 ## Proposed Changes Updates dependencies across the workspace. Any crate that was not able to be brought to the latest version is listed in #1712. ## Additional Info NA
2020-10-05 08:22:19 +00:00
serde = "1.0.116"
serde_derive = "1.0.116"
serde_json = "1.0.58"
Kintsugi review comments (#2831) * Fix makefile * Return on invalid finalized block * Fix todo in gossip scoring * Require --merge for --fee-recipient * Bump eth2_serde_utils * Change schema versions * Swap hash/uint256 test_random impls * Use default for ExecutionPayload::empty * Check for DBs before removing * Remove kintsugi docker image * Fix CLI default value
2021-11-29 22:32:53 +00:00
eth2_serde_utils = "0.1.1"
Enable Compatibility with Windows (#2333) ## Issue Addressed Windows incompatibility. ## Proposed Changes On windows, lighthouse needs to default to STDIN as tty doesn't exist. Also Windows uses ACLs for file permissions. So to mirror chmod 600, we will remove every entry in a file's ACL and add only a single SID that is an alias for the file owner. Beyond that, there were several changes made to different unit tests because windows has slightly different error messages as well as frustrating nuances around killing a process :/ ## Additional Info Tested on my Windows VM and it appears to work, also compiled & tested on Linux with these changes. Permissions look correct on both platforms now. Just waiting for my validator to activate on Prater so I can test running full validator client on windows. Co-authored-by: ethDreamer <37123614+ethDreamer@users.noreply.github.com> Co-authored-by: Michael Sproul <micsproul@gmail.com>
2021-05-19 23:05:16 +00:00
filesystem = { path = "../../common/filesystem" }
Fix assert in slashing protection import (#2881) ## Issue Addressed There was an overeager assert in the import of slashing protection data here: https://github.com/sigp/lighthouse/blob/fff01b24ddedcd54486e374460855ca20d3dd232/validator_client/slashing_protection/src/slashing_database.rs#L939 We were asserting that if the import contained any blocks for a validator, then the database should contain only a single block for that validator due to pruning/consolidation. However, we would only prune if the import contained _relevant blocks_ (that would actually change the maximum slot): https://github.com/sigp/lighthouse/blob/fff01b24ddedcd54486e374460855ca20d3dd232/validator_client/slashing_protection/src/slashing_database.rs#L629-L633 This lead to spurious failures (in the form of `ConsistencyError`s) when importing an interchange containing no new blocks for any of the validators. This wasn't hard to trigger, e.g. export and then immediately re-import the same file. ## Proposed Changes This PR fixes the issue by simplifying the import so that it's more like the import for attestations. I.e. we make the assert true by always pruning when the imported file contains blocks. In practice this doesn't have any downsides: if we import a new block then the behaviour is as before, except that we drop the `signing_root`. If we import an existing block or an old block then we prune the database to a single block. The only time this would be relevant is during extreme clock drift locally _plus_ import of a non-drifted interchange, which should occur infrequently. ## Additional Info I've also added `Arbitrary` implementations to the slashing protection types so that we can fuzz them. I have a fuzzer sitting in a separate directory which I may or may not commit in a subsequent PR. There's a new test in the standard interchange tests v5.2.1 that checks for this issue: https://github.com/eth-clients/slashing-protection-interchange-tests/pull/12
2022-01-04 20:46:44 +00:00
arbitrary = { version = "1.0", features = ["derive"], optional = true }
Implement Slashing Protection (#1116) * Implement slashing protection Roll-up of #588 with some conflicts resolved * WIP improvements * Require slot uniqueness for blocks (rather than epochs) * Native DB support for Slot and Epoch * Simplify surrounding/surrounded-by queries * Implement unified slashing protection database A single SQL database saves on open file descriptors. * Make slashing protection concurrency safe. Revive tests, add parallel tests. * Some simplifications * Auto-registration, test clean-ups * More tests, clean-ups, hardening * Fix comments in BLS * Optimise bulk validator registration * Delete outdated tests * Use bundled SQLite in slashing protection * Auto-register validators in simulation * Use real signing_root in slashing protection * Update book for --auto-register * Refine log messages and help flags * Correct typo in Cargo.toml authors * Fix merge conflicts * Safer error handling in sqlite slot/epoch * Address review comments * Add attestation test mutating block root Co-authored-by: pscott <scottpiriou@gmail.com>
2020-05-18 06:25:16 +00:00
[dev-dependencies]
Minify slashing protection interchange data (#2380) ## Issue Addressed Closes #2354 ## Proposed Changes Add a `minify` method to `slashing_protection::Interchange` that keeps only the maximum-epoch attestation and maximum-slot block for each validator. Specifically, `minify` constructs "synthetic" attestations (with no `signing_root`) containing the maximum source epoch _and_ the maximum target epoch from the input. This is equivalent to the `minify_synth` algorithm that I've formally verified in this repository: https://github.com/michaelsproul/slashing-proofs ## Additional Info Includes the JSON loading optimisation from #2347
2021-06-21 05:46:36 +00:00
lazy_static = "1.4.0"
Update external deps (#1711) ## Issue Addressed - Resolves #1706 ## Proposed Changes Updates dependencies across the workspace. Any crate that was not able to be brought to the latest version is listed in #1712. ## Additional Info NA
2020-10-05 08:22:19 +00:00
rayon = "1.4.1"
Fix assert in slashing protection import (#2881) ## Issue Addressed There was an overeager assert in the import of slashing protection data here: https://github.com/sigp/lighthouse/blob/fff01b24ddedcd54486e374460855ca20d3dd232/validator_client/slashing_protection/src/slashing_database.rs#L939 We were asserting that if the import contained any blocks for a validator, then the database should contain only a single block for that validator due to pruning/consolidation. However, we would only prune if the import contained _relevant blocks_ (that would actually change the maximum slot): https://github.com/sigp/lighthouse/blob/fff01b24ddedcd54486e374460855ca20d3dd232/validator_client/slashing_protection/src/slashing_database.rs#L629-L633 This lead to spurious failures (in the form of `ConsistencyError`s) when importing an interchange containing no new blocks for any of the validators. This wasn't hard to trigger, e.g. export and then immediately re-import the same file. ## Proposed Changes This PR fixes the issue by simplifying the import so that it's more like the import for attestations. I.e. we make the assert true by always pruning when the imported file contains blocks. In practice this doesn't have any downsides: if we import a new block then the behaviour is as before, except that we drop the `signing_root`. If we import an existing block or an old block then we prune the database to a single block. The only time this would be relevant is during extreme clock drift locally _plus_ import of a non-drifted interchange, which should occur infrequently. ## Additional Info I've also added `Arbitrary` implementations to the slashing protection types so that we can fuzz them. I have a fuzzer sitting in a separate directory which I may or may not commit in a subsequent PR. There's a new test in the standard interchange tests v5.2.1 that checks for this issue: https://github.com/eth-clients/slashing-protection-interchange-tests/pull/12
2022-01-04 20:46:44 +00:00
[features]
arbitrary-fuzz = ["arbitrary", "types/arbitrary-fuzz"]
Reference in New Issue Copy Permalink