GBHackers

Hackers Abuse Linux Cron Jobs to Maintain Persistent Malware Access on Firewalls and Switches


A Linux botnet targeting internet-facing edge devices can reinstall itself within five minutes if responders delete the malware but leave its scheduled job intact.

The campaign highlights a blind spot for operators of Linux-based firewalls and switches: reviewing the vendor configuration does not show whether the underlying operating system is running unauthorized jobs.

Evooo1Bot repurposes Mirai’s distributed denial-of-service engine and adds encrypted command-and-control and a SOCKS relay. Its persistence routine creates a cron job that retrieves and executes a remote script every five minutes.

If the scheduler remains active and the download server is reachable, deleting the binary buys responders only until the next scheduled run.

The bot can also turn a compromised appliance into a proxy, allowing operators to route traffic through the device or use it to reach internal networks.

Flare documented a different recovery mechanism in its February 9 analysis of SSHStalker, captured by an SSH honeypot.

Its cron job runs every minute and checks whether the bot process is still alive. If the process is gone, a local script relaunches it.

The distinction matters during containment: blocking access to a malware download server can disrupt remote retrieval, but it does not stop a local watchdog from restarting a bot already on disk.

FortiGuard Labs disclosed Evooo1Bot, on August 13, 2026, after observing exploitation attempts against edge devices.

Linux Cron Jobs Exploitation

Neither report establishes that the campaigns compromised specific switch models. There is, however, a documented firewall precedent: MITRE ATT&CK’s T1053.003 entry cites Operation MidnightEclipse, in which attackers used cron to retrieve payloads from attacker-controlled infrastructure after compromising Palo Alto Networks firewalls.

Cron is the mechanism that keeps code running after initial access, not the vulnerability that gets attackers onto the device.

Use a vendor-supported collection method to export system crontabs, per-user crontabs, and scheduled scripts where available.

Review vendor-native schedulers and event handlers separately; a clean Linux crontab does not clear those mechanisms.

Record what was inspected on each appliance and mark devices without supported shell access or equivalent telemetry as uninspected, not clean.

Prioritize jobs that download and immediately execute code, invoke scripts from temporary or memory-backed directories, or appeared after an unexplained administrative login.

A job that runs every minute is not proof of compromise, and download command does not identify Evooo1Bot access.

Compare the job with a known-good baseline for the same model and firmware, then correlate changes with centralized authentication logs and outbound connection records.

If investigators confirm a malicious job, isolate the appliance as operationally feasible and preserve the schedule, referenced scripts, file metadata, and relevant logs.

Turn off the job before terminating the process so the scheduler cannot restart it during cleanup.

Evooo1Bot also installs a systemd service labeled “Apache HTTPD Cache Manager,” an init script, an rc.local entry, and a login script. Removing its crontab entry leaves those other paths intact.

Before returning the device to service, close the initial access path and verify that neither the job nor unexpected outbound traffic returns.

Where that telemetry is missing, baseline scheduled jobs and monitor appliance egress rather than treating a quiet dashboard as evidence of a clean device.

Stops Cyber threats before impact with 21 min faster MTTR. Integrate ANYRUN’s Sandbox in your SOC.



Source link