fix(dashmate): local nodes missing envs after group reset by pshenmic · Pull Request #839 · dashpay/platform

pshenmic · 2023-03-21T13:11:03Z

Issue being fixed or feature implemented

After dashmate group reset, nodes fails to start due some contracts private key is missing

#785

What was done?

Moved contracts private keys generation in the separate factory and used it in the setup and reset phases

How Has This Been Tested?

Locally

Breaking Changes

No

Checklist:

I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
I have added or updated relevant unit/integration/functional/e2e tests
I have made corresponding changes to the documentation

For repository code-owners and collaborators only

I have assigned this pull request to a milestone

shumkov · 2023-03-22T02:03:17Z

+ * @param {generateHDPrivateKeys} generateHDPrivateKeys
+ * @return {configureTenderdashTask}
+ */
+function configureContractsPrivateKeysTaskFactory(generateHDPrivateKeys) {


generateSystemDataContractKeysTask

shumkov · 2023-03-22T02:03:35Z

 config.set('platform.featureFlags', baseConfig.platform.featureFlags);
 config.set('platform.masternodeRewardShares', baseConfig.platform.masternodeRewardShares);

+ subTasks.unshift({


…up-reset # Conflicts: # packages/dashmate/src/commands/group/reset.js # packages/dashmate/src/createDIContainer.js # packages/dashmate/src/listr/tasks/setup/setupLocalPresetTaskFactory.js

shumkov · 2023-04-19T06:36:20Z

+ // eslint-disable-next-line no-param-reassign
+ task.output = `DPNS Private Key: ${dpnsPrivateKey.toString()}`;
+
+ // eslint-disable-next-line no-param-reassign
+ task.output = `Feature Flags Private Key: ${featureFlagsPrivateKey.toString()}`;
+
+ // eslint-disable-next-line no-param-reassign
+ task.output = `Dashpay Private Key: ${dashpayPrivateKey.toString()}`;
+
+ // eslint-disable-next-line no-param-reassign
+ task.output = `Masternode Reward Shares Private Key: ${masternodeRewardSharesPrivateKey.toString()}`;


I think we don't need this anymore

shumkov · 2023-04-19T06:37:58Z

 const subTasks = ctx.configGroup.map((config, i) => (
 {
 title: `Create ${config.getName()} config`,
+ // eslint-disable-next-line consistent-return


shumkov · 2023-04-19T06:39:21Z

+ task: () => generateSystemDataContractKeysTask(config, config.get('network')),
+ },
+ {
+ task: () => resetNodeTask(config),


Didn't we reset config there?

shumkov · 2023-04-19T17:24:50Z

+ task: () => resetNodeTask(config),
+ },
+ {
+ task: () => generateSystemDataContractKeysTask(config, config.get('network')),


Did you test it? I think it won't work because all configs will have different keys now so init chain won't be deterministic. If you see how it worked before. We generated it once on setup and set to all configs. You should do the same

pshenmic · 2023-05-01T07:10:09Z

Fixed in #991

Bring the branch up to v4.1-dev, which bumps rust-dashcore to afcff156: key-wallet now impls `Zeroize` + `Drop` on `ExtendedPrivKey`, and `extended_public_key` moved into the new `ExtendedPubKeySigner` subtrait. Conflicts resolved (take base's rust-dashcore, afcff156): - Cargo.toml / Cargo.lock: base's afcff156 rev. - packages/rs-sdk-ffi/src/mnemonic_resolver_core_signer.rs: reconcile the derive-sign-destroy zeroization with the upstream change. - `WipingXprv` REMOVED — it was a local workaround for `ExtendedPrivKey` having no `Drop`, which key-wallet now provides (dashcore PR #833, added specifically for this signer). The bare value self-wipes on every path. - `WipingSecretKey` KEPT — the `secp256k1::SecretKey` copy at the two sign sites is the one key intermediate #833 cannot cover (`SecretKey` is an upstream type with no `Zeroize`, only `non_secure_erase()`), so the RAII guard remains the only way to close its `?`-early-return / panic leak window. Preserves the "no key bytes survive the trait boundary" invariant. - `extended_public_key` moved into an `impl ExtendedPubKeySigner` block (dashcore #839). Bump ripple + follow-through: - packages/rs-platform-wallet-ffi/src/dashpay.rs: import `ExtendedPubKeySigner` for `receiving_xpub` (the method moved off `Signer`). - packages/rs-platform-wallet-ffi/src/sign_with_mnemonic_resolver.rs: drop its own now-redundant `WipingXprv`, relying on `ExtendedPrivKey`'s upstream Drop. rs-sdk-ffi + platform-wallet-ffi build; 300 + 8 tests pass (incl. the 12 resolver-signer tests); full-workspace build green; fmt + clippy clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…endedPubKeySigner The v4.1-dev merge bumped rust-dashcore to afcff156, which (dashcore #839) moved `extended_public_key` out of `key_wallet::signer::Signer` into the new `ExtendedPubKeySigner: Signer` subtrait. Production callers were migrated in the merge, but four test-code signer impls still declared the method on `Signer` / `KwSigner`, so the workspace test build failed with `error[E0407]: method extended_public_key is not a member of trait ...` (dpp + platform-wallet lib tests). A plain `cargo build --workspace` compiles no test targets, which is why the merge's build looked clean; `cargo check --workspace --all-targets` surfaces it. Move the method into a separate `impl ExtendedPubKeySigner` block for each test signer (FixedKeySigner ×3, SeedSigner ×1); bodies are byte-for-byte identical (`ExtendedPubKeySigner: Signer` inherits the associated Error), so this is a pure trait-membership move, no behavior change. `cargo check --workspace --all-targets` clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

grovedb #839 (squashed as 6c882c3e) retires the standalone indexed-axis provers/verifiers, leaving PathQuery as the only public proof surface. This PR already routes everything through PathQuery, so the bump is a pin move from develop~1 (6b34ea81) with no platform code changes. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…prover retirement grovedb #836 gave the unified unproved read the attested skip (`PathQueryRun::AxisKeys/AxisEntries { .., skipped }`), and #839 retired the standalone indexed-axis provers, verifiers and read entry points outright — PathQuery is grovedb's only public surface for the family. The pin moves to that head and platform completes the unification it was waiting on: - Single-prefix unproved reads (ranked top-k and having-range) are one `run_path_query` over `PathQuery::new_axis` with the keys-only projection; the per-axis triple dispatches are gone, and the rank attestation maps from the new `skipped` field (its absence on a paginated read is corrupted state, not a default). - Single-prefix proofs are `prove_query(new_axis_top_k / new_axis_bounded)`; since transactional proving no longer exists anywhere, the committed-state-only guard covers single-prefix and branched proves alike. - Single-prefix verification matches `VerifiedPathQuery::AxisEntries { root_hash, entries, skipped }` through the same shared mappers as the branched arms. - Every read, proof and verification on both surfaces now builds exactly one PathQuery per external query; rs-drive re-exports `grovedb_query` so downstream crates can construct axis queries. Two semantic upgrades ride the retirement, both now pinned by tests: - A bound matching nothing against an EMPTY secondary proves: the envelope commits the element's empty secondary, authenticating complete absence instead of refusing ("Cannot create proof for empty tree" is gone from this surface; drive-abci's InvalidArgument mapping for that class is vestigial here and kept as dead-defensive). - The limit binds by RECONSTRUCTION, not echo: the verifier re-executes the proof under the queried limit, so an exhausted-walk proof is a complete answer under any cap that admits it (the old echo check rejected that soundly-verifiable case), while the dangerous direction — a proof truncated by a smaller limit verified under a larger cap — is rejected for missing coverage. The tamper test now pins BOTH directions; it previously exercised only the benign one. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

fix(dashmate): generate contract keys after group reset
c8aa7a5

pshenmic requested review from antouhou and shumkov as code owners March 21, 2023 13:11

pshenmic changed the base branch from v0.24-dev to masterMarch 21, 2023 13:13

shumkov requested changes Mar 22, 2023
View reviewed changes

pshenmic changed the base branch from master to v0.24-devApril 18, 2023 19:05

pshenmic added 2 commits April 18, 2023 22:11

Merge remote-tracking branch 'origin/v0.24-dev' into fix/dashmate-gro…
46ef622
…up-reset # Conflicts: # packages/dashmate/src/commands/group/reset.js # packages/dashmate/src/createDIContainer.js # packages/dashmate/src/listr/tasks/setup/setupLocalPresetTaskFactory.js

chore(dashmate): lint and bug fixes
9413ccf

shumkov requested changes Apr 19, 2023
View reviewed changes

chore(dashmate): code review fixes
95160bc

shumkov requested changes Apr 19, 2023
View reviewed changes

fix(dashmate): leave contract keys on group reset
4e4080f

pshenmic marked this pull request as draft April 20, 2023 14:11

pshenmic added 3 commits April 20, 2023 18:52

fix(dashmate): print system contracts private keys
9bea33d

fix(dashmate): revert generateSystemDataContractKeysTask
1cfac2e

chore(dashmate): revert script
6c8f23e

pshenmic closed this May 1, 2023

shumkov deleted the fix/dashmate-group-reset branch August 25, 2024 08:51

thepastaclaw mentioned this pull request Jul 2, 2026
chore: bump rust-dashcore to afcff156, export xpub via ExtendedPubKeySigner #3976
Merged
5 tasks

Navigation Menu

Uh oh!

fix(dashmate): local nodes missing envs after group reset - #839

fix(dashmate): local nodes missing envs after group reset#839
pshenmic wants to merge 8 commits into
v0.24-devfrom
fix/dashmate-group-reset

pshenmic commented Mar 21, 2023

Uh oh!

shumkovMar 22, 2023

Uh oh!

shumkovMar 22, 2023

Uh oh!

shumkovApr 19, 2023

Uh oh!

shumkovApr 19, 2023

Uh oh!

shumkovApr 19, 2023

Uh oh!

shumkovApr 19, 2023

Uh oh!

pshenmic commented May 1, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

pshenmic commented Mar 21, 2023

Issue being fixed or feature implemented

What was done?

How Has This Been Tested?

Breaking Changes

Checklist:

Uh oh!

shumkovMar 22, 2023

Choose a reason for hiding this comment

Uh oh!

shumkovMar 22, 2023

Choose a reason for hiding this comment

Uh oh!

shumkovApr 19, 2023

Choose a reason for hiding this comment

Uh oh!

shumkovApr 19, 2023

Choose a reason for hiding this comment

Uh oh!

shumkovApr 19, 2023

Choose a reason for hiding this comment

Uh oh!

shumkovApr 19, 2023

Choose a reason for hiding this comment

Uh oh!

pshenmic commented May 1, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants