GHSA-VH64-54PX-QGF8
Vulnerability from github – Published: 2025-03-03 16:21 – Updated: 2025-03-11 17:17Goroutine Leak in Abacus SSE Implementation
Summary
A critical goroutine leak vulnerability has been identified in the Abacus server's Server-Sent Events (SSE) implementation. The issue occurs when clients disconnect from the /stream endpoint, as the server fails to properly clean up resources and terminate associated goroutines. This leads to resource exhaustion where the server continues running but eventually stops accepting new SSE connections while maintaining high memory usage. The vulnerability specifically involves improper channel cleanup in the event handling mechanism, causing goroutines to remain blocked indefinitely.
POC
Impact
This vulnerability affects all versions of Abacus prior to v1.4.0. The issue causes:
- Permanent unresponsiveness of the
/streamendpoint after prolonged use - Memory growth that stabilizes at a high level but prevents proper functionality
- Selective denial of service affecting only SSE connections while other endpoints remain functional
- Accumulated orphaned goroutines that cannot be garbage collected
- High resource consumption under sustained client connection/disconnection patterns
Systems running Abacus in production with client applications that frequently establish and terminate SSE connections are most vulnerable. The issue becomes particularly apparent in high-traffic environments or during connection stress testing.
Patches
The vulnerability has been patched in Abacus v1.4.0. The fix includes:
- Implementing buffered channels to prevent blocking operations during cleanup
- Adding proper mutex-protected cleanup logic to ensure resources are released exactly once
- Implementing timeout protection for channel operations to prevent deadlocks
- Ensuring consistent cleanup when connections terminate unexpectedly
- Adding improved monitoring for client disconnections using request context
- Restructuring the event broadcasting system to safely handle client removal
Users should upgrade to v1.4.0 or later versions as soon as possible.
Workarounds
If upgrading is not immediately possible, the following workarounds can help mitigate the issue:
-
Limit maximum connections: Configure your reverse proxy to limit the maximum number of concurrent connections to the
/streamendpoints. -
Implement request timeouts: Configure your infrastructure to terminate long-lived SSE connections after a reasonable period.
-
Restart regularly: Schedule regular restarts of the Abacus service to reclaim resources.
-
Monitor memory usage: Set up alerts for abnormal memory growth patterns.
-
Separate instance for SSE: Run a dedicated Abacus instance solely for handling SSE connections, allowing it to be restarted without affecting the main API functionality.
References
- Go Concurrency Patterns: Context
- CWE-772: Missing Release of Resource after Effective Lifetime
- OWASP Top 10: Resource Exhaustion
- Resource Management in Go
For More Information
Please contact the Abacus security team at abacus@jsn.cam for additional information or to report further security issues.
{
"affected": [
{
"package": {
"ecosystem": "Go",
"name": "github.com/jasonlovesdoggo/abacus"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.0.0-20250302043802-898ff1204e11"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2025-27421"
],
"database_specific": {
"cwe_ids": [
"CWE-400",
"CWE-772"
],
"github_reviewed": true,
"github_reviewed_at": "2025-03-03T16:21:46Z",
"nvd_published_at": "2025-03-03T17:15:15Z",
"severity": "HIGH"
},
"details": "## Goroutine Leak in Abacus SSE Implementation\n\n### Summary\n\nA critical goroutine leak vulnerability has been identified in the Abacus server\u0027s Server-Sent Events (SSE) implementation. The issue occurs when clients disconnect from the `/stream` endpoint, as the server fails to properly clean up resources and terminate associated goroutines. This leads to resource exhaustion where the server continues running but eventually stops accepting new SSE connections while maintaining high memory usage. The vulnerability specifically involves improper channel cleanup in the event handling mechanism, causing goroutines to remain blocked indefinitely.\n\n### [POC](https://github.com/JasonLovesDoggo/abacus/blob/main/docs/bugs/GHSA-vh64-54px-qgf8/test.py)\n\n### Impact\n\nThis vulnerability affects all versions of Abacus prior to v1.4.0. The issue causes:\n\n- Permanent unresponsiveness of the `/stream` endpoint after prolonged use\n- Memory growth that stabilizes at a high level but prevents proper functionality\n- Selective denial of service affecting only SSE connections while other endpoints remain functional\n- Accumulated orphaned goroutines that cannot be garbage collected\n- High resource consumption under sustained client connection/disconnection patterns\n\nSystems running Abacus in production with client applications that frequently establish and terminate SSE connections are most vulnerable. The issue becomes particularly apparent in high-traffic environments or during connection stress testing.\n\n### Patches\n\nThe vulnerability has been patched in Abacus v1.4.0. The fix includes:\n\n1. Implementing buffered channels to prevent blocking operations during cleanup\n2. Adding proper mutex-protected cleanup logic to ensure resources are released exactly once\n3. Implementing timeout protection for channel operations to prevent deadlocks\n4. Ensuring consistent cleanup when connections terminate unexpectedly\n5. Adding improved monitoring for client disconnections using request context\n6. Restructuring the event broadcasting system to safely handle client removal\n\nUsers should upgrade to v1.4.0 or later versions as soon as possible.\n\n### Workarounds\n\nIf upgrading is not immediately possible, the following workarounds can help mitigate the issue:\n\n1. **Limit maximum connections**: Configure your reverse proxy to limit the maximum number of concurrent connections to the `/stream` endpoints.\n\n2. **Implement request timeouts**: Configure your infrastructure to terminate long-lived SSE connections after a reasonable period.\n\n3. **Restart regularly**: Schedule regular restarts of the Abacus service to reclaim resources.\n\n4. **Monitor memory usage**: Set up alerts for abnormal memory growth patterns.\n\n5. **Separate instance for SSE**: Run a dedicated Abacus instance solely for handling SSE connections, allowing it to be restarted without affecting the main API functionality.\n\n### References\n\n- [Go Concurrency Patterns: Context](https://blog.golang.org/context)\n- [CWE-772: Missing Release of Resource after Effective Lifetime](https://cwe.mitre.org/data/definitions/772.html)\n- [OWASP Top 10: Resource Exhaustion](https://owasp.org/www-project-top-ten/2017/A5_2017-Broken_Access_Control)\n- [Resource Management in Go](https://go.dev/blog/defer-panic-and-recover)\n\n### For More Information\n\nPlease contact the Abacus security team at abacus@jsn.cam for additional information or to report further security issues.",
"id": "GHSA-vh64-54px-qgf8",
"modified": "2025-03-11T17:17:30Z",
"published": "2025-03-03T16:21:46Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/JasonLovesDoggo/abacus/security/advisories/GHSA-vh64-54px-qgf8"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2025-27421"
},
{
"type": "WEB",
"url": "https://github.com/JasonLovesDoggo/abacus/commit/78fdb9b48b7f6d08ed0cd41077509c0a97071552"
},
{
"type": "WEB",
"url": "https://github.com/JasonLovesDoggo/abacus/commit/898ff1204e11317cc161240b660e63eed5a72b33"
},
{
"type": "PACKAGE",
"url": "https://github.com/JasonLovesDoggo/abacus"
},
{
"type": "WEB",
"url": "https://pkg.go.dev/vuln/GO-2025-3498"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
}
],
"summary": "Goroutine Leak in Abacus SSE Implementation"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.