feat: add script to delete data older than 90 days#2693
Open
wilsonrivera wants to merge 6 commits intomainfrom
Open
feat: add script to delete data older than 90 days#2693wilsonrivera wants to merge 6 commits intomainfrom
wilsonrivera wants to merge 6 commits intomainfrom
Conversation
wilsonrivera
requested review from
JivusAyrus,
StarpTech and
thisisnithin
as code owners
Contributor
WalkthroughAdds five PostgreSQL indexes via a migration and journal entry, updates Drizzle schema declarations to declare those indexes, and introduces a new TypeScript cleanup script that deletes old rows from several tables in batched transactions. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes 🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (5)
controlplane/src/bin/cleanup-old-data.ts (5)
105-145: Add explicit return type annotation.Per TypeScript coding guidelines, functions should have explicit return type annotations.
Proposed fix
async function deleteOldDataInChunks( db: PostgresJsDatabase<typeof schema>, table: | typeof schema.schemaChecks | typeof schema.graphCompositions | typeof schema.auditLogs | typeof schema.webhookDeliveries, -) { +): Promise<number> {As per coding guidelines: "Use explicit type annotations for function parameters and return types in TypeScript".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 105 - 145, The function deleteOldDataInChunks is missing an explicit return type; add a TypeScript return annotation so the signature declares it returns Promise<number> (keep the existing parameters like db: PostgresJsDatabase<typeof schema> and the union table type unchanged) to satisfy the project's explicit type annotation guideline.
178-234: Add explicit return type annotation.Proposed fix
-async function deleteOldSchemaVersions(db: PostgresJsDatabase<typeof schema>) { +async function deleteOldSchemaVersions(db: PostgresJsDatabase<typeof schema>): Promise<number> {As per coding guidelines: "Use explicit type annotations for function parameters and return types in TypeScript".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 178 - 234, The function deleteOldSchemaVersions is missing an explicit return type; add an explicit Promise<number> return annotation to its signature (the db parameter can remain typed as PostgresJsDatabase<typeof schema>) so the declaration becomes deleteOldSchemaVersions(db: PostgresJsDatabase<typeof schema>): Promise<number>, ensuring the async function's return type is explicit and matches the returned deleteCount number.
12-14: Consider making the retention period configurable.The 90-day cutoff is hardcoded. Consider making this configurable via environment variable or command-line argument for operational flexibility.
Proposed approach
+const DEFAULT_RETENTION_DAYS = 90; +const RETENTION_DAYS = Number(process.env.CLEANUP_RETENTION_DAYS) || DEFAULT_RETENTION_DAYS; + -const CUTOFF_DATE = startOfDay(subDays(new Date(), 90)); +const CUTOFF_DATE = startOfDay(subDays(new Date(), RETENTION_DAYS));🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 12 - 14, CUTOFF_DATE is hardcoded to 90 days; make the retention period configurable by reading a RETENTION_DAYS environment variable or a command-line flag and fall back to 90 if not provided/invalid. Update the code that defines CUTOFF_DATE to parse and validate the variable (e.g., Number(process.env.RETENTION_DAYS) or a parsed CLI arg) and then compute startOfDay(subDays(new Date(), retentionDays)); leave ITEMS_PER_CHUNK and NUMBER_OF_PARALLEL_TRANSACTIONS as-is but consider adding similar env/CLI overrides if desired.
32-94: Consider adding a dry-run mode for safety.For a destructive operation like this, a
--dry-runflag that logs what would be deleted without actually deleting would help operators validate the cleanup scope before execution, especially in production environments.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 32 - 94, Add a --dry-run option to the cleanup script so the process logs what would be removed without performing deletes: parse process.argv for a dryRun boolean at script startup and thread it into the deletion calls used in this file (pass to deleteOldDataInChunks and deleteOldSchemaVersions), then modify those functions to accept a dryRun flag and short-circuit actual DELETE operations — instead run the same SELECT/count logic and log the IDs/counts and messages using CUTOFF_DATE and the existing console messages; keep the same timing and final summary but ensure no DB mutations happen when dryRun is true and still close queryConnection in finally.
161-176: Add explicit return type annotation.Proposed fix
-async function getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>) { +async function getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>): Promise<string[]> {As per coding guidelines: "Use explicit type annotations for function parameters and return types in TypeScript".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 161 - 176, Add an explicit return type to the getFederatedGraphsFlags function signature (e.g. Promise<Array<number | string>> or the precise id type used by your schema) so the function signature declares its return type instead of relying on inference; update the signature for getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>): Promise<Array<number | string>> (or replace number | string with the exact type of baseCompositionSchemaVersionId/composedSchemaVersionId from schema) to satisfy the coding guideline.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 105-145: The function deleteOldDataInChunks is missing an explicit
return type; add a TypeScript return annotation so the signature declares it
returns Promise<number> (keep the existing parameters like db:
PostgresJsDatabase<typeof schema> and the union table type unchanged) to satisfy
the project's explicit type annotation guideline.
- Around line 178-234: The function deleteOldSchemaVersions is missing an
explicit return type; add an explicit Promise<number> return annotation to its
signature (the db parameter can remain typed as PostgresJsDatabase<typeof
schema>) so the declaration becomes deleteOldSchemaVersions(db:
PostgresJsDatabase<typeof schema>): Promise<number>, ensuring the async
function's return type is explicit and matches the returned deleteCount number.
- Around line 12-14: CUTOFF_DATE is hardcoded to 90 days; make the retention
period configurable by reading a RETENTION_DAYS environment variable or a
command-line flag and fall back to 90 if not provided/invalid. Update the code
that defines CUTOFF_DATE to parse and validate the variable (e.g.,
Number(process.env.RETENTION_DAYS) or a parsed CLI arg) and then compute
startOfDay(subDays(new Date(), retentionDays)); leave ITEMS_PER_CHUNK and
NUMBER_OF_PARALLEL_TRANSACTIONS as-is but consider adding similar env/CLI
overrides if desired.
- Around line 32-94: Add a --dry-run option to the cleanup script so the process
logs what would be removed without performing deletes: parse process.argv for a
dryRun boolean at script startup and thread it into the deletion calls used in
this file (pass to deleteOldDataInChunks and deleteOldSchemaVersions), then
modify those functions to accept a dryRun flag and short-circuit actual DELETE
operations — instead run the same SELECT/count logic and log the IDs/counts and
messages using CUTOFF_DATE and the existing console messages; keep the same
timing and final summary but ensure no DB mutations happen when dryRun is true
and still close queryConnection in finally.
- Around line 161-176: Add an explicit return type to the
getFederatedGraphsFlags function signature (e.g. Promise<Array<number | string>>
or the precise id type used by your schema) so the function signature declares
its return type instead of relying on inference; update the signature for
getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>):
Promise<Array<number | string>> (or replace number | string with the exact type
of baseCompositionSchemaVersionId/composedSchemaVersionId from schema) to
satisfy the coding guideline.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: a25c1271-e9ba-410c-8f4e-5e72d87b484a
📒 Files selected for processing (6)
controlplane/migrations/0137_certain_monster_badoon.sqlcontrolplane/migrations/meta/0137_snapshot.jsoncontrolplane/migrations/meta/_journal.jsoncontrolplane/package.jsoncontrolplane/src/bin/cleanup-old-data.tscontrolplane/src/db/schema.ts
Codecov Report❌ Patch coverage is
❌ Your patch check has failed because the patch coverage (2.27%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage. Additional details and impacted files@@ Coverage Diff @@
## main #2693 +/- ##
==========================================
+ Coverage 63.10% 64.13% +1.02%
==========================================
Files 245 306 +61
Lines 26273 43633 +17360
Branches 0 4666 +4666
==========================================
+ Hits 16580 27983 +11403
- Misses 8352 15627 +7275
+ Partials 1341 23 -1318
🚀 New features to boost your workflow:
|
JivusAyrus
requested changes
Mar 27, 2026
| // Delete compositions | ||
| deleteStart = performance.now(); | ||
| console.log(`Deleting compositions created before ${CUTOFF_DATE.toDateString()}...`); | ||
| const compositionsDeleted = await deleteOldDataInChunks(db, schema.graphCompositions); |
Member
There was a problem hiding this comment.
why delete graph compositions? on deletion of schema versions the graph compositions get deleted right?
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
controlplane/src/bin/cleanup-old-data.ts (1)
107-114: Add explicit return types to the remaining helpers.
deleteOldDataInChunks,getFederatedGraphsFlags, anddeleteOldCompositionsstill rely on inference. Please declare theirPromise<number>/Promise<string[]>return types so this file matches the repo TypeScript rule.As per coding guidelines, "Use explicit type annotations for function parameters and return types in TypeScript".
Also applies to: 163-163, 180-180
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 107 - 114, Add explicit return type annotations to the helper functions to satisfy the repo TypeScript rule: declare deleteOldDataInChunks as returning Promise<number>, getFederatedGraphsFlags as returning Promise<string[]>, and deleteOldCompositions as returning Promise<number>. Update the function signatures (for the symbols deleteOldDataInChunks, getFederatedGraphsFlags, deleteOldCompositions) to include these return types while leaving the internal logic unchanged.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 247-248: The current logic pushes schemaVersionIds from
graphCompositions and graphCompositionSubgraphs into the global
idsOfSchemaVersionsToKeep via getSchemaVersionIds (the two all.push lines),
which incorrectly treats every old composition as a version to retain and
prevents deleteOldSchemaVersions()/deleteOldCompositions() from cleaning up;
change the logic so the keep-set is built only from rows you actually intend to
retain (e.g., only compositions marked active/retained) or maintain separate
keep-sets (e.g., compositionSchemaVersionIds vs retainedSchemaVersionIds) and
use the appropriate set when calling deleteOldSchemaVersions and
deleteOldCompositions; locate and update the code paths around
getSchemaVersionIds, idsOfSchemaVersionsToKeep, deleteOldSchemaVersions, and
deleteOldCompositions to either filter
graphCompositions/graphCompositionSubgraphs rows by intended-retention criteria
or split them into distinct keep sets.
- Around line 183-190: The code currently selects all expired graphCompositions
into memory (variable graphCompositions) — change this to page through results
using a loop that re-queries with a limit and an orderBy (e.g., order by
createdAt and use a BATCH_SIZE) and delete each page before fetching the next,
following the pattern in deleteOldDataInChunks(); apply the same
pagination/fetch-by-page fix for the schemaVersion selection block referenced in
the comment (the block at ~228-232) so no full retention set is materialized in
memory and each iteration re-queries with lte(schemaX.createdAt, CUTOFF_DATE),
orderBy createdAt and limit BATCH_SIZE until no rows remain.
---
Nitpick comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 107-114: Add explicit return type annotations to the helper
functions to satisfy the repo TypeScript rule: declare deleteOldDataInChunks as
returning Promise<number>, getFederatedGraphsFlags as returning
Promise<string[]>, and deleteOldCompositions as returning Promise<number>.
Update the function signatures (for the symbols deleteOldDataInChunks,
getFederatedGraphsFlags, deleteOldCompositions) to include these return types
while leaving the internal logic unchanged.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 9d402453-a2ee-46bd-a570-4a87854eeb1d
📒 Files selected for processing (1)
controlplane/src/bin/cleanup-old-data.ts
Comment on lines
+183
to
+190
| const graphCompositions = await db | ||
| .select({ | ||
| id: schema.graphCompositions.id, | ||
| schemaVersionId: schema.graphCompositions.schemaVersionId, | ||
| }) | ||
| .from(schema.graphCompositions) | ||
| .where(lte(schema.graphCompositions.createdAt, CUTOFF_DATE)) | ||
| .execute(); |
Contributor
There was a problem hiding this comment.
Page these candidates instead of loading the full retention set into memory.
Both helpers materialize every expired graphCompositions / schemaVersion row before batching deletes. On a retention run, that can create a large memory spike and a long initial scan. Re-query each batch with limit/orderBy, like deleteOldDataInChunks() does.
Also applies to: 228-232
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@controlplane/src/bin/cleanup-old-data.ts` around lines 183 - 190, The code
currently selects all expired graphCompositions into memory (variable
graphCompositions) — change this to page through results using a loop that
re-queries with a limit and an orderBy (e.g., order by createdAt and use a
BATCH_SIZE) and delete each page before fetching the next, following the pattern
in deleteOldDataInChunks(); apply the same pagination/fetch-by-page fix for the
schemaVersion selection block referenced in the comment (the block at ~228-232)
so no full retention set is materialized in memory and each iteration re-queries
with lte(schemaX.createdAt, CUTOFF_DATE), orderBy createdAt and limit BATCH_SIZE
until no rows remain.
JivusAyrus
requested changes
Apr 1, 2026
| console.log(` ${schemaVersionsDeleted} schema versions deleted in ${deleteDurationInSeconds.toFixed(3)} seconds`); | ||
|
|
||
| // Delete compositions | ||
| if (schemaVersionsDeleted > 0) { |
Member
There was a problem hiding this comment.
if the graph compositions are getting deleted after the schema versions, then we need not do them at all right? as they get cascaded on deleting schema versions
Contributor
Author
There was a problem hiding this comment.
You are correct, I was overcomplicating it
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
♻️ Duplicate comments (2)
controlplane/src/bin/cleanup-old-data.ts (2)
192-193:⚠️ Potential issue | 🟠 MajorComposition-derived IDs in the keep-set block old schema-version cleanup.
Line 192-193 adds
graphCompositions/graphCompositionSubgraphsschemaVersion IDs to the global keep-set, which can keep old schema versions alive solely because old compositions still reference them. That defeats the intended cascade-based cleanup path for old compositions.💡 Suggested fix
- all.push(...(await getSchemaVersionIds(db, schema.graphCompositions, 'schemaVersionId'))); - all.push(...(await getSchemaVersionIds(db, schema.graphCompositionSubgraphs, 'schemaVersionId')));🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 192 - 193, The keep-set population is incorrectly adding schemaVersion IDs derived from compositions which prevents cascade cleanup; remove the two pushes that call getSchemaVersionIds for schema.graphCompositions and schema.graphCompositionSubgraphs (the lines referencing schema.graphCompositions and schema.graphCompositionSubgraphs and getSchemaVersionIds) so composition-derived schemaVersionIds are not added to the global keep-set and allow compositions to be cleaned up via the intended cascade path.
173-177:⚠️ Potential issue | 🟠 MajorAvoid materializing all expired schema versions in memory.
Line 173-177 loads the entire expired
schemaVersionset before chunking withsplice(Line 201-205). On large retention runs, this can create a big memory spike and long initial scan; page candidates in a loop (limit/order/cursor) instead.Also applies to: 201-205
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 173 - 177, The code currently materializes all expired schema versions into schemaVersions via db.select(...).from(schema.schemaVersion).where(lte(schema.schemaVersion.createdAt, CUTOFF_DATE)).execute(), then chunks with splice which causes large memory spikes; change this to a paginated query loop that repeatedly fetches a bounded page (e.g., limit N) ordered by schema.schemaVersion.createdAt (and tie-breaker id), process each page, and advance via a cursor/last-seen key until no more rows, and remove the splice-based chunking logic (references: schemaVersions, schema.schemaVersion, CUTOFF_DATE). Ensure each page is processed and deleted before fetching the next to keep memory constant.
🧹 Nitpick comments (1)
controlplane/src/bin/cleanup-old-data.ts (1)
97-104: Add explicit return types to async helper functions.
deleteOldDataInChunksandgetFederatedGraphsFlagscurrently rely on inferred return types. Please annotate them explicitly for consistency and safer API contracts.💡 Suggested fix
async function deleteOldDataInChunks( db: PostgresJsDatabase<typeof schema>, table: | typeof schema.schemaChecks | typeof schema.graphCompositions | typeof schema.auditLogs | typeof schema.webhookDeliveries, -) { +): Promise<number> { @@ async function getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>) { +async function getFederatedGraphsFlags(db: PostgresJsDatabase<typeof schema>): Promise<string[]> {As per coding guidelines, "
**/*.{ts,tsx}: Use explicit type annotations for function parameters and return types in TypeScript."Also applies to: 153-153
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@controlplane/src/bin/cleanup-old-data.ts` around lines 97 - 104, Annotate the async helper functions with explicit return types: add ": Promise<void>" to the deleteOldDataInChunks function declaration so its contract is explicit, and add an explicit Promise return type to getFederatedGraphsFlags (e.g. ": Promise<Record<string, boolean>>" or the appropriate typed interface if it returns a more specific shape) to replace inferred types; update the function signatures for deleteOldDataInChunks and getFederatedGraphsFlags accordingly so TypeScript enforces the return contract.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 49-60: The cleanup deletes schema versions via
deleteOldSchemaVersions (returning idsOfSchemaVersionsToKeep) before removing
audit logs, which can violate the FK on audit_logs.schemaVersionId and abort;
reorder or adjust logic so audit logs are deleted before schema versions (call
deleteOldDataInChunks for schema.auditLogs prior to invoking
deleteOldSchemaVersions), or alternatively ensure deleteOldSchemaVersions uses
audit log data to populate idsOfSchemaVersionsToKeep (include schemaVersionIds
referenced by old audit logs) so referenced schema versions are retained; update
the sequence around deleteOldSchemaVersions, deleteOldDataInChunks, and the
idsOfSchemaVersionsToKeep handling to enforce FK-safe deletion.
---
Duplicate comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 192-193: The keep-set population is incorrectly adding
schemaVersion IDs derived from compositions which prevents cascade cleanup;
remove the two pushes that call getSchemaVersionIds for schema.graphCompositions
and schema.graphCompositionSubgraphs (the lines referencing
schema.graphCompositions and schema.graphCompositionSubgraphs and
getSchemaVersionIds) so composition-derived schemaVersionIds are not added to
the global keep-set and allow compositions to be cleaned up via the intended
cascade path.
- Around line 173-177: The code currently materializes all expired schema
versions into schemaVersions via
db.select(...).from(schema.schemaVersion).where(lte(schema.schemaVersion.createdAt,
CUTOFF_DATE)).execute(), then chunks with splice which causes large memory
spikes; change this to a paginated query loop that repeatedly fetches a bounded
page (e.g., limit N) ordered by schema.schemaVersion.createdAt (and tie-breaker
id), process each page, and advance via a cursor/last-seen key until no more
rows, and remove the splice-based chunking logic (references: schemaVersions,
schema.schemaVersion, CUTOFF_DATE). Ensure each page is processed and deleted
before fetching the next to keep memory constant.
---
Nitpick comments:
In `@controlplane/src/bin/cleanup-old-data.ts`:
- Around line 97-104: Annotate the async helper functions with explicit return
types: add ": Promise<void>" to the deleteOldDataInChunks function declaration
so its contract is explicit, and add an explicit Promise return type to
getFederatedGraphsFlags (e.g. ": Promise<Record<string, boolean>>" or the
appropriate typed interface if it returns a more specific shape) to replace
inferred types; update the function signatures for deleteOldDataInChunks and
getFederatedGraphsFlags accordingly so TypeScript enforces the return contract.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: a9ed15ce-c8fc-40a1-885e-1c7f2d0ba137
📒 Files selected for processing (1)
controlplane/src/bin/cleanup-old-data.ts
Comment on lines
+49
to
+60
| // Delete schema versions | ||
| deleteStart = performance.now(); | ||
| console.log(`Deleting schema versions created before ${CUTOFF_DATE.toDateString()}...`); | ||
| const [schemaVersionsDeleted, idsOfSchemaVersionsToKeep] = await deleteOldSchemaVersions(db); | ||
|
|
||
| deleteDurationInSeconds = (performance.now() - deleteStart) / 1000; | ||
| console.log(` ${schemaVersionsDeleted} schema versions deleted in ${deleteDurationInSeconds.toFixed(3)} seconds`); | ||
|
|
||
| // Delete audit logs | ||
| deleteStart = performance.now(); | ||
| console.log(`Deleting audit logs created before ${CUTOFF_DATE.toDateString()}...`); | ||
| const auditLogsDeleted = await deleteOldDataInChunks(db, schema.auditLogs); |
Contributor
There was a problem hiding this comment.
Delete order can violate FK constraints and stop the cleanup job.
At Line 52, schema versions are deleted before audit logs (Line 60), but audit_logs.schemaVersionId has onDelete: 'no action' (controlplane/src/db/schema.ts:2030-2040). Since audit-log schemaVersion IDs are not included in the keep-set (Line 187-194), a schema version referenced only by old audit logs can fail deletion and abort the script.
💡 Suggested fix
- // Delete schema versions
- deleteStart = performance.now();
- console.log(`Deleting schema versions created before ${CUTOFF_DATE.toDateString()}...`);
- const [schemaVersionsDeleted, idsOfSchemaVersionsToKeep] = await deleteOldSchemaVersions(db);
-
- deleteDurationInSeconds = (performance.now() - deleteStart) / 1000;
- console.log(` ${schemaVersionsDeleted} schema versions deleted in ${deleteDurationInSeconds.toFixed(3)} seconds`);
-
// Delete audit logs
deleteStart = performance.now();
console.log(`Deleting audit logs created before ${CUTOFF_DATE.toDateString()}...`);
const auditLogsDeleted = await deleteOldDataInChunks(db, schema.auditLogs);
deleteDurationInSeconds = (performance.now() - deleteStart) / 1000;
console.log(` ${auditLogsDeleted} audit logs deleted in ${deleteDurationInSeconds.toFixed(3)} seconds`);
+
+ // Delete schema versions
+ deleteStart = performance.now();
+ console.log(`Deleting schema versions created before ${CUTOFF_DATE.toDateString()}...`);
+ const [schemaVersionsDeleted, idsOfSchemaVersionsToKeep] = await deleteOldSchemaVersions(db);
+
+ deleteDurationInSeconds = (performance.now() - deleteStart) / 1000;
+ console.log(` ${schemaVersionsDeleted} schema versions deleted in ${deleteDurationInSeconds.toFixed(3)} seconds`); all.push(...(await getFederatedGraphsFlags(db)));
+ all.push(...(await getSchemaVersionIds(db, schema.auditLogs, 'schemaVersionId')));
all.push(...(await getSchemaVersionIds(db, schema.pluginImageVersions, 'schemaVersionId')));Also applies to: 183-194
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@controlplane/src/bin/cleanup-old-data.ts` around lines 49 - 60, The cleanup
deletes schema versions via deleteOldSchemaVersions (returning
idsOfSchemaVersionsToKeep) before removing audit logs, which can violate the FK
on audit_logs.schemaVersionId and abort; reorder or adjust logic so audit logs
are deleted before schema versions (call deleteOldDataInChunks for
schema.auditLogs prior to invoking deleteOldSchemaVersions), or alternatively
ensure deleteOldSchemaVersions uses audit log data to populate
idsOfSchemaVersionsToKeep (include schemaVersionIds referenced by old audit
logs) so referenced schema versions are retained; update the sequence around
deleteOldSchemaVersions, deleteOldDataInChunks, and the
idsOfSchemaVersionsToKeep handling to enforce FK-safe deletion.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Summary by CodeRabbit
Checklist
Open Source AI Manifesto
This project follows the principles of the Open Source AI Manifesto. Please ensure your contribution aligns with its principles.