Skip to main content
Vulnerability Database/CVE-2026-68498

CVE-2026-68498: Jackson-core DOS Vulnerability

CVE-2026-68498 is a denial of service flaw in Jackson-core that allows attackers to exhaust heap memory through oversized JSON property names. This post explains its impact, affected versions, and mitigation steps.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-68498 Overview

CVE-2026-68498 is a resource exhaustion vulnerability in the Jackson-core JSON library. The flaw affects com.fasterxml.jackson.core:jackson-core from version 2.16.0 and tools.jackson.core:jackson-core from version 3.0.0. The ReaderBasedJsonParser fails to enforce StreamReadConstraints.maxNameLength incrementally when parsing JSON Object property names from character-based input. Instead of bounding allocation at the documented ~50,000 characters, the parser allows names to grow up to maxStringLength (100,000,000 characters by default). A remote, unauthenticated attacker can submit JSON with a single oversized property name to force ~200 MB allocations per request and exhaust the Java heap.

Critical Impact

Unauthenticated remote attackers can trigger heap exhaustion and crash Java services that parse untrusted JSON from String, Reader, or char[] sources.

Affected Products

  • com.fasterxml.jackson.core:jackson-core versions 2.16.0 through 2.18.9
  • tools.jackson.core:jackson-core version 3.0.0
  • Any application parsing untrusted JSON via ReaderBasedJsonParser (character input sources)

Discovery Timeline

  • 2026-10-16 - CVE-2026-68498 published to the National Vulnerability Database
  • 2026-10-20 - Last updated in NVD database

Technical Details for CVE-2026-68498

Vulnerability Analysis

The vulnerability resides in the character-input slow path of ReaderBasedJsonParser._parseName2(). This method accumulates JSON Object property names into a TextBuffer. When a buffer segment fills, the parser calls TextBuffer.finishCurrentSegment(), which validates only against maxStringLength — a limit intended for string values, not property names.

The correct boundary check, StreamReadConstraints.validateNameLength(int), runs later inside CharsToNameCanonicalizer.findSymbol(). That validation executes only after the entire name has been scanned to its closing quote. By then, the parser has already buffered up to 100,000,000 characters in memory, roughly 2000 times the maxNameLength default of 50,000. This classifies as [CWE-400] Uncontrolled Resource Consumption.

Byte-input parsers including UTF8StreamJsonParser, UTF8DataInputJsonParser, and the non-blocking async parsers are not affected. These implementations revalidate maxNameLength on every growth of the name-decode buffer via ParserBase._growNameDecodeBuffer().

Root Cause

The root cause is a missing incremental length check in the character-input parsing path. The parser reuses the string-value validation logic for property name accumulation, which applies the wrong constraint. Because StreamReadConstraints.maxDocumentLength is disabled by default (-1), no outer bound stops the oversized token before memory is consumed.

Attack Vector

An attacker sends an HTTP request containing a JSON body with a single property name of several million characters. If the receiving application reads the body into a String, Reader, or char[] before parsing — common in Spring MVC and similar frameworks — Jackson routes parsing through the vulnerable ReaderBasedJsonParser. Each concurrent request allocates approximately 200 MB. A small number of parallel requests can exhaust the heap and crash the JVM.

text
#1643: Enforce maxNameLength incrementally in ReaderBasedJsonParser [GHSA-649p-m576-vr99]
 (fix by @tinyb0y)

2.18.9 (07-Jul-2026)

Source: GitHub Jackson Core Commit 2cd206de. The patch adds incremental enforcement of maxNameLength inside ReaderBasedJsonParser so oversized names are rejected before significant allocation.

Detection Methods for CVE-2026-68498

Indicators of Compromise

  • Repeated OutOfMemoryError: Java heap space exceptions in application logs tied to JSON parsing stack frames
  • Stack traces referencing ReaderBasedJsonParser._parseName2 or TextBuffer.finishCurrentSegment immediately before heap failures
  • Unexpected spikes in HTTP request body sizes to JSON endpoints followed by JVM crashes or container restarts

Detection Strategies

  • Inventory running Java services for Jackson-core versions between 2.16.0 and 2.18.9 and all 3.0.0 builds using software composition analysis tools
  • Inspect application code for parsing patterns that pass String, Reader, or char[] sources to ObjectMapper.readValue() rather than raw byte[] or InputStream
  • Enable JVM heap dump on OutOfMemoryError and review dumps for large char[] arrays owned by TextBuffer

Monitoring Recommendations

  • Alert on sudden JVM heap exhaustion events, garbage collection pressure, and container OOMKill signals on services exposing JSON APIs
  • Deploy WAF rules that cap request body size and reject JSON payloads containing property name tokens above a reasonable threshold
  • Correlate HTTP 5xx surges and application restarts with inbound payload size metrics to identify active abuse

How to Mitigate CVE-2026-68498

Immediate Actions Required

  • Upgrade com.fasterxml.jackson.core:jackson-core to version 2.18.10 or later across all services
  • Audit frameworks and libraries that transitively pull Jackson-core and force the fixed version in build dependency management
  • Where upgrades are blocked, switch JSON parsing to byte-input sources (InputStream or byte[]) which are not affected

Patch Information

The fix is tracked in GitHub Jackson Core Pull Request #1643 and delivered in commit 2cd206de. It is included in jackson-core 2.18.10. Full details are published in GitHub Security Advisory GHSA-649p-m576-vr99. Review the Jackson Core 2.x release notes and 3.x release notes for version-specific guidance.

Workarounds

  • Configure StreamReadConstraints.maxDocumentLength to a reasonable byte ceiling to bound total token input before parsing begins
  • Terminate TLS and parse JSON at an upstream gateway that enforces strict body-size and token-size limits before reaching Java services
  • Refactor controllers to accept raw request bodies as InputStream rather than String, routing parsing through the unaffected UTF8StreamJsonParser
bash
# Maven dependency pin to a patched release
<dependency>
  <groupId>com.fasterxml.jackson.core</groupId>
  <artifactId>jackson-core</artifactId>
  <version>2.18.10</version>
</dependency>

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.