can1357/oh-my-pi · error · ArchiveError

Archive is too large to read in memory (${formatBytes(size)}

Error message

Archive is too large to read in memory (${formatBytes(size)} > ${formatBytes(limits.maxInMemorySize)} limit)

What it means

assertInMemorySize rejects archives that would be fully loaded into memory beyond the configured limit (default maxInMemorySize is 256 MiB in DEFAULT_ARCHIVE_LIMITS). The message reports both the measured size and the limit via formatBytes so you can see by how much the archive overshoots.

Source

Thrown at packages/utils/src/ar/limits.ts:41

	maxLinkDepth: number;
}

export const DEFAULT_ARCHIVE_LIMITS: ArchiveLimits = {
	maxEntries: 1_000_000,
	maxInMemorySize: 256 * 1024 * 1024,
	maxIndexSize: 64 * 1024 * 1024,
	maxMemberSize: 64 * 1024 * 1024,
	maxPathBytes: 4096,
	maxLinkDepth: 40,
};

/** Reject an archive that would be fully materialized beyond `maxInMemorySize`. */
export function assertInMemorySize(size: number, limits: ArchiveLimits): void {
	if (!Number.isSafeInteger(size) || size < 0) {
		throw new ArchiveError("Archive is too large to read safely");
	}
	if (size > limits.maxInMemorySize) {
		throw new ArchiveError(
			`Archive is too large to read in memory (${formatBytes(size)} > ${formatBytes(limits.maxInMemorySize)} limit)`,
		);
	}
}

/** Reject archive metadata (index/header) beyond `maxIndexSize`. */
export function assertIndexSize(size: number, limits: ArchiveLimits, what: string): void {
	if (!Number.isSafeInteger(size) || size < 0) {
		throw new ArchiveError(`Invalid archive: ${what} has an invalid size`);
	}
	if (size > limits.maxIndexSize) {
		throw new ArchiveError(
			`Archive ${what} is too large (${formatBytes(size)} > ${formatBytes(limits.maxIndexSize)} limit)`,
		);
	}
}

/**

View on GitHub (pinned to 9690622007)

Solutions

1. Pass custom limits raising maxInMemorySize for this call: { limits: { ...DEFAULT_ARCHIVE_LIMITS, maxInMemorySize: 2 * 1024 * 1024 * 1024 } } (mind actual available RAM)
2. Stream/extract the archive with a native tool (cpio, cabextract, dpkg-deb) instead of materializing it in JS
3. Split the archive or read only needed members via targeted tooling
4. Check for decompression bombs if the size seemed unexpectedly large; reject untrusted inputs

Example fix

// before
const entries = await readCpio(bigBuffer); // >256MiB -> ArchiveError
// after
import { DEFAULT_ARCHIVE_LIMITS } from '.../ar/limits';
const entries = await readCpio(bigBuffer, {
  limits: { ...DEFAULT_ARCHIVE_LIMITS, maxInMemorySize: 1024 * 1024 * 1024 },
});

import { stat } from 'node:fs/promises';
import { DEFAULT_ARCHIVE_LIMITS } from '.../ar/limits';
const { size } = await stat(path);
if (size > DEFAULT_ARCHIVE_LIMITS.maxInMemorySize) {
  throw new Error(`archive ${size} bytes exceeds in-memory limit; stream instead`);
}

null

try {
  entries = await readCpio(source);
} catch (err) {
  if (err instanceof ArchiveError && err.message.startsWith('Archive is too large to read in memory')) {
    const m = err.message.match(/\((.*?) > (.*?) limit\)/); // parse size vs limit
    // raise limits explicitly or switch to native streaming tools
  } else throw err;
}

When it happens

Trigger: Calling a materializing reader (readArj, readCabArchive, readCpio, readDeb -> decompressDebTar, etc.) with default limits on an archive (or a decompressed deb control/data tar member) larger than limits.maxInMemorySize — e.g. a 500 MiB cpio or a deb whose data.tar decompresses past 256 MiB.

Common situations: Reading large system images/backup archives (big cpio/cab), .deb packages with huge data payloads, CI machines where callers assumed smaller artifacts, or decompression bombs whose inflated size crosses the limit.

Related errors

• Archive is too large to read safely
• ${name} exceeds the ${SHARPSHOOTER_MAX_FILE_LINES}-line limi
• Invalid ARJ basic header size
• Invalid ARJ archive: too many extended headers
• ASAR payload is too large to encode safely

AI-assisted analysis of can1357/oh-my-pi@9690622007 (2026-08-31). Data as JSON: /api/errors/b632aa18e91f4ce5. Report an issue: GitHub.