Performance Optimization

ZinTrust is built for speed, but there are several techniques you can use to further optimize your application.

Memoization

Use Memoize.create(...) to cache the results of expensive function calls.

import { Memoize } from '@zintrust/core';

const getGlobalStats = async (): Promise\<unknown> => {
  // Expensive database aggregation
  return { ok: true };
};

export const getGlobalStatsMemoized = Memoize.create(getGlobalStats, { ttl: 60_000 });

Lazy Loading

ZinTrust supports lazy loading for heavy dependencies to improve startup time.

import { LazyLoader } from '@zintrust/core';

const loader = LazyLoader.create();
const bcrypt = await loader.load('bcrypt');

Parallel Execution

When performing multiple independent operations, use ParallelGenerator to run them concurrently.

import { ParallelGenerator } from '@zintrust/core';

await ParallelGenerator.runAll([() => fetchUser(), () => fetchPosts(), () => fetchSettings()]);

Database Optimization

• Eager Loading: Use with() to avoid N+1 queries.
• Indexing: Ensure your database columns are properly indexed.
• Query Caching: Cache frequent query results in Redis or memory.

Performance Optimization

ZinTrust is built for speed, but there are several techniques you can use to further optimize your application.

This page focuses on the performance utilities that are part of the public @zintrust/core API.

Memoization

Use Memoize.create(...) to cache the results of expensive function calls.

import { Memoize } from '@zintrust/core';

const getGlobalStats = async (): Promise\<unknown> => {
  // Expensive database aggregation
  return { ok: true };
};

export const getGlobalStatsMemoized = Memoize.create(getGlobalStats, {
  ttl: 60_000,
});

Notes:

• Memoization is in-memory per process (not shared across instances).
• By default the cache key is derived from JSON.stringify(args); prefer a custom keyGenerator when args are large or non-serializable.

Lazy Loading

Use LazyLoader to defer loading optional or heavy dependencies until they’re needed.

import { LazyLoader } from '@zintrust/core';

const loader = LazyLoader.create();
const bcrypt = await loader.load\<typeof import('bcrypt')>('bcrypt');

Notes:

• LazyLoader uses dynamic import() and caches the module namespace object.
• This is most useful for optional dependencies and rarely-executed code paths.

Parallel Execution

When performing multiple independent operations, use ParallelGenerator to run them concurrently.

import { ParallelGenerator } from '@zintrust/core';

await ParallelGenerator.runAll([() => fetchUser(), () => fetchPosts(), () => fetchSettings()]);

If you want to bound concurrency, use batching:

import { ParallelGenerator } from '@zintrust/core';

await ParallelGenerator.runBatch(
  [() => fetchUser(), () => fetchPosts(), () => fetchSettings(), () => fetchTeams()],
  2
);

Notes:

• runBatch() executes batches sequentially, and each batch runs in parallel.

PerformanceOptimizer

PerformanceOptimizer wraps caching, lazy-loading, and parallel execution behind a single interface.

import { PerformanceOptimizer } from '@zintrust/core';

const optimizer = PerformanceOptimizer.create();

const result = await optimizer.generateWithCache('stats', { tenantId: 't_123' }, async () => ({
  ok: true,
}));

await optimizer.preloadModules(['bcrypt']);

const parallelResults = await optimizer.generateInParallel(
  [() => Promise.resolve(1), () => Promise.resolve(2)],
  2
);

void result;
void parallelResults;

Notes:

• The optimizer’s generation cache persists to disk by default under .gen-cache/ in the current working directory.

Database Optimization

• Eager Loading: Use with() to avoid N+1 queries.
• Indexing: Ensure your database columns are properly indexed.
• Query Caching: Cache frequent query results in Redis or memory.

Production Mode

Always run ZinTrust in production mode (APP_ENV=production) to disable debug logging and enable production optimizations.