inferwire
/
Cybersecurity·2 min read

Exploited Zimbra Flaw Allows Command Execution; Fixed in 10.1.20

CVE-2026-73570 lets unauthenticated attackers run commands on Zimbra servers before 10.1.20 that have SNMP notifications enabled. CERT Polska and CISA both report active exploitation.

TL;DR

  • CVE-2026-73570 lets unauthenticated attackers run commands on Zimbra Collaboration before 10.1.20 when the optional zimbra-snmp package and SNMP notifications are enabled [^1].
  • CERT Polska warned of an ongoing campaign on August 17, and CISA added the flaw to its catalog of exploited vulnerabilities on August 21 [^2] [^1].

Background

Zimbra Collaboration is mail and groupware server software that organizations run on their own infrastructure. SNMP is a protocol for monitoring servers and network devices, and Zimbra's optional SNMP package can send notifications, called traps, to a monitoring system [^2]. In a command injection flaw, text that an attacker controls ends up being executed as a shell command.

What happened

According to NVD, Zimbra Collaboration (ZCS) before 10.1.20 doesn't properly sanitize untrusted input while processing SNMP notifications. An unauthenticated attacker can send specially crafted SMTP requests, the protocol used to deliver mail, which may result in arbitrary operating system commands running as the Zimbra user [^1]. NVD rates it 8.9, high, and classifies it as OS command injection (CWE-78) [^1]. Zimbra's list of security advisories records the fix in version 10.1.20 and credits the security operations center of the Vatican City State Governorate with the report [^3].

CERT Polska published its warning about a campaign exploiting the flaw on August 17 [^2]. It says the flaw affects instances where the SNMP trap service is enabled through the snmp_notify parameter and the swatchdog service is running, which it is by default [^2]. CISA added CVE-2026-73570 to its Known Exploited Vulnerabilities catalog on August 21 and gave US federal agencies until August 24 to apply mitigations or stop using the product [^1].

Why it matters

Mail servers accept SMTP traffic from the whole internet by design. A flaw that can be triggered with crafted SMTP requests and no login is therefore within reach of anyone who can send mail to the server, as long as the vulnerable configuration is in place [^1]. An attacker who gets in lands on the system that handles an organization's email. Because exploitation is already under way, CERT Polska advises administrators not only to update but also to look for signs that their server was attacked [^2].

Practical example

Say you administer a Zimbra server at version 10.1.19 with the zimbra-snmp package installed. After updating to 10.1.20, you work through CERT Polska's checklist [^2]. You search /var/log/zimbra.log for "Service status change" entries where an unexpected payload stands in for the service name, followed by "changed from stopped to running" or "changed from running to stopped" [^2]. Then you look for files the zimbra user created in the last 30 days in /opt/zimbra/jetty/webapps/, /opt/zimbra/jetty_base/webapps/ and /tmp/ [^2]. If either check turns up signs of exploitation, CERT Polska asks to be contacted right away [^2].

Sources

  1. [1]NVD — CVE-2026-73570
  2. [2]CERT Polska — Aktywnie wykorzystywana podatność w Zimbra Collaboration Suite
  3. [3]Zimbra — Zimbra Security Advisories