fix(satp-hermes): migrate all DBs on startup by mskrzypkows · Pull Request #4378 · hyperledger-cacti/cacti

mskrzypkows · 2026-06-10T14:41:58Z

The CLI only forwarded localRepository and remoteRepository to SATPGatewayConfig, leaving auditRepository and oracleLogRepository to fall back to a random UUID-named SQLite file. Those databases were never migrated, causing the gateway to crash on the first write to the audit or oracle log tables.

Validate and pass auditRepository and oracleLogRepository from the config JSON in the CLI, consistent with how local/remote repos are already handled
Extend createDBRepository() to run migrate.latest() across all four configured repositories instead of localRepository only
Add integration test asserting that startup() creates the expected tables (logs, remote-logs, audit_entries, oracle_logs) in all four databases

Copilot

Pull request overview

This PR fixes SATP Hermes gateway startup failures caused by un-migrated databases by ensuring all configured repositories (local, remote, audit, oracle) are validated, passed through the CLI, and migrated during gateway startup.

Changes:

CLI now validates and forwards auditRepository and oracleLogRepository from the JSON config into SATPGatewayConfig.
Gateway startup migration logic is extended from “local only” to include all configured repositories.
Integration coverage is added to assert that DB tables are created after startup() across the configured repositories.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
`packages/cactus-plugin-satp-hermes/src/test/typescript/integration/gateway/gateway-init-startup.test.ts`	Adds an integration test to verify that DB migrations create the expected tables on startup.
`packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway.ts`	Updates `createDBRepository()` to migrate more than just the local repository.
`packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway-cli.ts`	Validates and passes audit/oracle repository configs so they don’t fall back to unintended defaults.

AndreAugusto11 · 2026-06-11T11:08:52Z

@@ -1257,15 +1257,25 @@ export class SATPGateway implements IPluginWebService, ICactusPlugin {
        this.logger.info(
          `${fnTag}: Created migration source: ${JSON.stringify(migrationSource)}`,
        );


@mskrzypkows could you please address this comment as well?

Yes, I have a fix for it. Testing it and will push.

Based on the comment create better solution, calling repository.migrate for each repository from SATPGateway. This way we don't need additional connection to the DB.

Thank you @mskrzypkows. Looking better! Can you please make sure that:

errors initializing each DB or its migration are propagated up and handled

can you please add the DB cleanup to the onShutdown hook on the SATP gateway?

… on startup The CLI only forwarded localRepository and remoteRepository to SATPGatewayConfig, leaving auditRepository and oracleLogRepository to fall back to a random UUID-named SQLite file. Those databases were never migrated, causing the gateway to crash on the first write to the audit or oracle log tables. - Validate and pass auditRepository and oracleLogRepository from the config JSON in the CLI, consistent with how local/remote repos are already handled - Extend createDBRepository() to run migrate.latest() across all four configured repositories instead of localRepository only - Add integration test asserting that startup() creates the expected tables (logs, remote-logs, audit_entries, oracle_logs) in all four databases Assisted-by: anthropic:claude-sonnet-4.6 Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

createDBRepository() created temporary Knex instances from config to run migrate.latest(), then destroyed them. This caused three bugs: - For SQLite :memory: databases each Knex() call opens a separate in-memory DB, so migrations ran on a throwaway connection while the repository's own connection stayed unmigrated. - database.destroy() was not in a finally block, leaking connections when migrate.latest() threw. - auditRepository had no default fallback, unlike localRepository. Add migrate() to IRepository and implement it in all four concrete repositories so createDBRepository() migrates the actual runtime connections directly. Assisted-by: anthropic:claude-sonnet-4.6 Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

for the IPFSRemoteLogRepository Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

db_init_at_startup_pr Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

RafaelAPB · 2026-07-01T15:28:38Z

+          this.remoteRepository,
+          this.auditRepository,
+          this.oracleLogRepository,
+        ].filter((r): r is NonNullable<typeof r> => r != null);


Can you make the condition more explicit (on the filtering do the check if the database is non null and valid) instead of doing a non nullable type enforcement?

I moved explicit check into the loop.

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

RafaelAPB

@mskrzypkows I mean what is on the suggestions rest LGTM

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

mskrzypkows · 2026-07-06T21:39:43Z

Combined suggestions into one commit and added warning log.

RafaelAPB · 2026-07-07T22:48:04Z

@mskrzypkows thank you once again. I think we are good to go. Please 1) rebase with main; 2) squash your commits into one.

mskrzypkows requested review from AndreAugusto11, LordKubaya, RafaelAPB, VRamakrishna, outSH and sandeepnRES as code owners June 10, 2026 14:41

RafaelAPB requested a review from Copilot June 10, 2026 16:38

Copilot started reviewing on behalf of RafaelAPB June 10, 2026 16:38 View session

RafaelAPB reviewed Jun 10, 2026

View reviewed changes

Comment thread packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway.ts Outdated

Copilot AI reviewed Jun 10, 2026

View reviewed changes

RafaelAPB reviewed Jun 11, 2026

View reviewed changes

Comment thread packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway.ts Outdated

mskrzypkows and others added 6 commits June 12, 2026 13:39

fix(satp-hermes): use built-in function for checking table existence

257dfc7

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

fix(satp-hermes): missing migrate implementation

17e86dd

for the IPFSRemoteLogRepository Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

fix(satp-hermes): removed non-null operator usage

61665e1

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

fix(satp-hermes): shutdown hook for created repositories

8e0461d

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

mskrzypkows force-pushed the db_init_at_startup_pr branch from f9869c8 to 8e0461d Compare June 12, 2026 11:41

Merge branch 'main' of https://github.com/hyperledger/cacti into

684170f

db_init_at_startup_pr Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

mskrzypkows force-pushed the db_init_at_startup_pr branch from 7dbfd6b to 684170f Compare June 26, 2026 16:29

RafaelAPB requested changes Jul 1, 2026

View reviewed changes

fix(satp-hermes): address review feedback on DB migration startup

4966de5

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

RafaelAPB requested changes Jul 5, 2026

View reviewed changes

Comment thread packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway.ts Outdated

Comment thread packages/cactus-plugin-satp-hermes/src/main/typescript/plugin-satp-hermes-gateway.ts Outdated

fix(satp-hermes): explicit null check for shutdown callback

4cab7cd

Signed-off-by: Maciej Skrzypkowski <mskr@gmx.com>

Uh oh!

Conversation

mskrzypkows commented Jun 10, 2026

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

AndreAugusto11 Jun 11, 2026

Choose a reason for hiding this comment

Uh oh!

mskrzypkows Jun 11, 2026

Choose a reason for hiding this comment

Uh oh!

mskrzypkows Jun 11, 2026

Choose a reason for hiding this comment

Uh oh!

RafaelAPB Jun 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mskrzypkows Jun 12, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

RafaelAPB Jul 1, 2026

Choose a reason for hiding this comment

Uh oh!

mskrzypkows Jul 4, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

RafaelAPB left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

mskrzypkows commented Jul 6, 2026

Uh oh!

RafaelAPB commented Jul 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

RafaelAPB Jun 11, 2026 •

edited

Loading