a9f3c50863
The previous LLM-generated knowledge files contained factual
hallucinations. The most visible was the claim that `FindFirst` /
`FindLast` "forces a full-table scan" on an unfiltered record - it does
not; those APIs return a single row via the current key.
Other inaccuracies the audit found and fixed:
* `FindSet(true)` was described as "taking a LockTable". The correct
upstream phrasing is that `FindSet(true)` sets
`ReadIsolation::UpdLock` on the read. UpdLock and LockTable are
related but distinct mechanisms.
* The list of production-scale tables had been invented beyond the
upstream source (e.g. "Detailed Cust. Ledg. Entry") without a
citation. The regenerated list matches the ten tables upstream lists
with their P95 row counts.
* `SetLoadFields` guidance had been augmented with an extra mechanism
claim ("the database resolves the filter using the index without
hydrating the value") not present in upstream.
Approach: full regeneration of `microsoft/knowledge/` from the six
upstream BCApps Code Review instruction files, with Microsoft Learn /
the AL language reference as a secondary source. Every claim in every
regenerated file is anchored to a verbatim upstream quote (or a Learn
URL); the audit trail lives in artifacts/trace-<domain>.json on the
session workspace.
The PR #11 transaction/error-handling cluster is preserved verbatim:
* performance/understand-implicit-transaction-boundary.md
* performance/codeunit-run-as-atomic-sub-operation.{md,good.al,bad.al}
* performance/codeunit-run-requires-prior-commit-inside-transaction.{md,good.al,bad.al}
* performance/use-tryfunction-for-error-catching-not-rollback.{md,good.al,bad.al}
* performance/avoid-commit-inside-loops.{md,good.al,bad.al}
* security/commitbehavior-attribute-scopes-explicit-commits.{md,good.al,bad.al}
* testing/transactionmodel-attribute-governs-test-transactions.{md,good.al,bad.al}
These articles already cite Microsoft Learn and were carefully
cross-referenced; the regeneration skips their topics rather than
duplicating them.
File counts after regeneration:
performance 35 .md (5 preserved + 30 new)
privacy 17 .md
security 18 .md (1 preserved + 17 new)
style 33 .md
testing 1 .md (preserved)
ui 19 .md
upgrade 18 .md
Total 141 atomic knowledge files, each strictly one rule. All pass
.github/scripts/validate_frontmatter.py with 0 errors and 0 warnings.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
1.9 KiB
| bc-version | domain | keywords | technologies | countries | application-area | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
performance |
|
|
|
|
Production-scale tables warrant concrete performance analysis
Description
Some Business Central tables routinely reach sizes where access patterns matter much more than they do on a generic table. The upstream review guidance lists ten of them with P95 row counts: Item (~800k), Customer (~800k), Item Ledger Entry (~10M), Value Entry (~10M), G/L Entry (~10M), VAT Entry (~10M), Customer Ledger Entry (~10M), Vendor Ledger Entry (~10M), Sales Invoice Header (~300k), and Sales Invoice Line (~3M). These figures are not platform constants — they are the volumes a reviewer should assume when judging a change.
Best Practice
For any code change that touches one of these tables, do not approve the pattern on intuition. Walk through the SQL the change implies (one query? one per row? one per chunk?), the memory it allocates (a List per row?), and the CPU work per row, against the row counts above. Smaller tables can tolerate a sub-optimal access pattern; these cannot. The rest of this domain — apply-filters-before-iterating.md, use-setloadfields-for-partial-records.md, avoid-calcfields-in-loops.md, pair-findset-with-next-loop.md, avoid-get-inside-loop-on-persistent-tables.md — exists primarily so that code touching these tables stays on the safe side of each rule.
Anti Pattern
Generalizing from a unit test or a development tenant. A FindSet loop with a per-row CalcFields may execute in milliseconds against a few thousand rows on a developer's machine and become a multi-minute table scan against ten million Value Entry rows in production. Reasoning about performance from the dev-tenant timing instead of the production volume is the single most common way a regression ships.