CVE-2021-34433 is a missing signature check (CWE-347) in the DTLS client of Eclipse Californium 2.0.0 to 2.6.4 and 3.0.0-M1 to 3.0.0-M3: a certificate-based handshake succeeds even when the server never signs its ServerKeyExchange. NVD scores it 7.5. Californium fixed it in 2.6.5 and 3.0.0-M4; Leshan users need 1.3.2 or 2.0.0-M5.
Vulnerability summary
CVSS base score
7.5
High severity
EPSS score (FIRST)
0.3%
Probability of exploitation in the wild in the next 30 days
CISA KEV catalog
No
Not listed
Published
Vulnerability ID: CVE-2021-34433 · Source: NVD
Affected components
| Component | CPE | Version range |
|---|---|---|
| Eclipse Californium (Scandium DTLS) | cpe:2.3:a:eclipse:californium:*:*:*:*:*:*:*:* | 2.0.0 to 2.6.4 and 3.0.0-M1 to 3.0.0-M3 (NVD, CVE record); fixed in 2.6.5 and 3.0.0-M4 (Californium release notes) |
| Eclipse Leshan (LwM2M, bundles Californium) | — | 1.0.0 to 1.3.1 with X.509 or RPK; fixed in 1.3.2 (Californium 2.6.6) and 2.0.0-M5 (Californium 3.0.0-RC1) (Leshan release notes) |
What does this mean for affected devices?
The description, written by the Eclipse Foundation as CNA, reads: "In Eclipse Californium version 2.0.0 to 2.6.4 and 3.0.0-M1 to 3.0.0-M3, the certificate based (x509 and RPK) DTLS handshakes accidentally succeeds without verifying the server side's signature on the client side, if that signature is not included in the server's ServerKeyExchange." NVD's Primary score is 7.5, CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:N, a high integrity impact with no confidentiality or availability impact. NVD classifies it as CWE-347 (improper verification of a cryptographic signature); the CNA lists CWE-322 (key exchange without entity authentication) and gives no score. The record was published on 2021-08-20 (NVD, CVE record).
Why does a missing signature matter?
In a certificate or raw public key handshake, the ServerKeyExchange signature is what proves the server holds the private key behind the certificate it presented. A Californium client that accepts the message without that signature has not authenticated the server, so a peer able to sit in the path can present a certificate it does not own. Pre-shared key handshakes are outside the CVE description; Californium's release notes address the fix to users "which are using DTLS RPK/x509 for authentication".
Which versions fix it?
Californium's 3.0.0-M4 release says the bugfix "is required for all users of Californium 3.0.0-M1 - 3.0.0-M3" using DTLS RPK or X.509, and the 2.6.5 release says the same for 2.0.0 to 2.6.4. Californium's security policy lists the flaw as a "DTLS certificates verification bypass" and names 2.6.6 among the 2.x bugfix releases.
Which devices are affected?
Californium is a Java CoAP and DTLS stack, so it runs on gateways and on the server side of device fleets rather than on microcontrollers. Eclipse Leshan, the project's LwM2M device management library, bundles it: the Leshan 1.3.2 release integrates Californium 2.6.6 and says people using X.509 or RPK with Leshan 1.0.0 to 1.3.1 "MUST upgrade", and Leshan 2.0.0-M5 integrates Californium 3.0.0-RC1 with the fix. Check the scandium artifact version in every gateway image and LwM2M server. Our CoAP security guide covers the DTLS modes. The later Californium flaw CVE-2022-39368 affects the same DTLS layer, and CVE-2020-3162 is a CoAP flaw in Cisco's field network manager.
CVE-2021-34433 is not in the CISA KEV catalog as of 2026-11-18 (catalog version 2026.11.17). FIRST EPSS gives it 0.00338, the 27.7th percentile, in the 2026-11-17 dataset (CISA KEV, FIRST EPSS).
Evidence
- NVD: Californium 2.0.0 to 2.6.4 and 3.0.0-M1 to 3.0.0-M3, certificate based (x509 and RPK) DTLS handshakes succeed without verifying the server signature if it is not included in ServerKeyExchange; [email protected] Primary CVSS 3.1 7.5 AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:N; CWE-347 (NVD), CWE-322 ([email protected]); CPE californium 2.0.0 to < 2.6.5 and 3.0.0-M1..M3; published 2021-08-20. — NVD
- Californium 3.0.0-M4 release notes: bugfix required for all users of 3.0.0-M1 - 3.0.0-M3 using DTLS RPK/x509 for authentication, provides the fix for CVE-2021-34433. — Eclipse Californium on GitHub
- Californium 2.6.5 release notes: bugfix release required for all users of 2.0.0 - 2.6.4 using DTLS RPK/x509 for authentication, provides the fix for CVE-2021-34433. — Eclipse Californium on GitHub
- Californium security policy: known vulnerability "DTLS certificates verification bypass CVE-2021-34433"; bugfix releases listed include 2.6.6. — Eclipse Californium security policy
- Leshan 1.3.2 (2021-12-17): integration of Californium 2.6.6 which fixes CVE-2021-34433; users of x509 and/or RPK with Leshan 1.0.0 to 1.3.1 must upgrade. Leshan 2.0.0-M5 (2021-10-21): integration of Californium 3.0.0-RC1 which fixes CVE-2021-34433. — Eclipse Leshan on GitHub
- CVE record (CNA Eclipse Foundation): Eclipse Californium 2.0.0 to 2.6.4 and 3.0.0-M1 to 3.0.0-M3 affected; reference Eclipse bug 575281. — CVE.org
- Not in the CISA KEV catalog, version 2026.11.17 (1,782 entries), checked 2026-11-18. — CISA KEV
- EPSS 0.00338, percentile 0.2768, dataset 2026-11-17, read 2026-11-18. — FIRST EPSS

Related advisories
Sources
Frequently asked questions
Which Californium versions fix CVE-2021-34433?
Californium 2.6.5 for the 2.x line and 3.0.0-M4 for the 3.0 milestones; later releases such as 2.6.6 and 3.0.0 include the fix (per the Californium release notes and security policy).
Are pre-shared key (PSK) handshakes affected by CVE-2021-34433?
The CVE describes certificate-based handshakes, X.509 and raw public key, and Californium addresses the fix to users of DTLS RPK or X.509 authentication. PSK handshakes are outside that description (per NVD and the Californium release notes).
Which Leshan versions are affected by CVE-2021-34433?
Leshan 1.0.0 to 1.3.1 when X.509 or RPK is used. Leshan 1.3.2 integrates Californium 2.6.6 and Leshan 2.0.0-M5 integrates Californium 3.0.0-RC1, both with the fix (per the Leshan release notes).
Is CVE-2021-34433 exploited in the wild?
It is not in the CISA KEV catalog as of 2026-11-18 (catalog version 2026.11.17). FIRST EPSS gives it 0.00338, the 27.7th percentile, in the 2026-11-17 dataset (per CISA KEV and FIRST EPSS).
Questions about this page?