Hackers Use Hidden Cron Jobs to Reinstall Malware on Firewalls and Routers Every 5 Minutes

Blog WriterCybersecurity News - Original News Source is cybersecuritynews.com


A Linux botnet is hiding in a place most network teams never look: the scheduled jobs on their firewalls, routers and switches.

If a responder kills the malware and deletes the file, a hidden cron job downloads it again within five minutes. The botnet is called Evooo1Bot. FortiGuard Labs first described it on August 13, 2026.

It goes after internet-facing devices such as routers, firewalls, IP cameras, phone systems and storage boxes.

The malware reuses the attack engine from the leaked Mirai source code, and its authors have added encrypted control traffic, a password-guessing scanner and a proxy feature.

Analysts at Eclypsium noted this weakness in a November 17 report. They warned that killing the process does not clean a compromised device when a scheduled job quietly puts the malware back. Many teams treat that step as the end of cleanup, and that is where this botnet wins.

Hackers Use Hidden Cron Jobs

Cron is the standard Linux job scheduler. It runs tasks at set times, and many network operating systems are built on Linux, so they include it. The legitimate jobs on a firewall are usually small. They rotate logs, clean up files and run health checks.

Evooo1Bot abuses this. One bot command, !persist, installs several ways to stay on the device at once. The cron entry runs every five minutes.

It uses wget or curl to download a script and pipes it straight into a shell. If the bot process is killed and the binary is deleted, the next five-minute mark pulls everything back.

The cron job is only one layer. The same command also creates a systemd service with the harmless-sounding name “Apache HTTPD Cache Manager,” an init script, an /etc/rc.local entry and a login script under /etc/profile.d/.

Removing the cron line alone leaves those in place. The malware also lowers its priority for the system’s out-of-memory killer and keeps the watchdog file open, so a hardware reboot is less likely to interrupt it.

Cron is not how the botnet gets in. According to FortiGuard, it spreads by exploiting known flaws in devices from D-Link, Tenda, NETGEAR, Telesquare and others, and by guessing SSH passwords.

All the exploit callbacks pointed to one loader address. Once inside, the bot can turn a device into a SOCKS proxy. FortiGuard called that the most important feature, because the attacker can hide behind your device and move deeper into your network.

This is not a one-off trick. Flare researchers described the SSHStalker botnet in February 2026. It uses a cron job that runs every minute and restarts the malware if it disappears.

Our earlier report on a legacy IRC botnet campaign showed how that one-minute watchdog brings bots back in about 60 seconds.

How Defenders Can Find and Remove Hidden Jobs

Appliances are easy to miss. Admins work in the vendor’s command line, and configuration tools track the vendor settings, not the Linux crontab underneath.

Endpoint agents usually do not run on these devices. As we have seen when hackers target edge devices, that blind spot is exactly why attackers like them.

Eclypsium advises starting in the device’s Linux shell. On Arista EOS, the bash command opens one. On Cisco NX-OS, the bash shell must be enabled first, and many deployments leave it off. If you cannot enable it, record that cron was not inspected. Do not record the device as clean.

Read the scheduler files directly. Check /etc/crontab, /etc/cron.d/, the hourly, daily, weekly and monthly folders, and the spool files under /var/spool/cron/.

Do not rely on crontab -l alone. An attacker with root can write straight into the spool folder without ever running that command. The spool path also differs by Linux family, so check both /var/spool/cron/crontabs/ and /var/spool/cron/.

Then search for risky patterns. A curl or wget command piped into sh or bash points to Evooo1Bot. References to /dev/tcp, nc or ncat suggest a reverse shell, and a long base64 string suggests a hidden command.

Jobs that run every minute or at boot deserve a closer look, but some vendors run their own frequent jobs, so treat those as suspicious and not proof. Any job that downloads and runs code should go to incident response.

Order matters during cleanup. Save a copy of the cron file and its timestamps, remove the cron entry first, and only then kill the process. Doing it the other way gives the scheduler another chance to run.

After that, check the other startup spots, and look at vendor schedulers such as NX-OS EEM applets or Arista event handlers, which are not cron. Eclypsium also warned that it has not tested whether a cron entry survives a reboot or upgrade on every platform.

Patching helps stop the first break-in, but it will not remove a job that is already there. In the stealthy PEERSTITCH malware implant case, attackers also relied on weak visibility inside firewalls.

Indicators of compromise (IoCs):-

Type Indicator Notes
IP address 91.92.40[.]118 Loader host seen in exploit callbacks
URL 91.92.40[.]118/wget.sh Loader script that downloads the bot
SHA-256 f13cb360768363d3424e2192c7805b8c8015eb8706dbbbcdead6aed8cf390109 Evooo1Bot sample
SHA-256 4c0886349e9d348569fffe1b7a31e474d514508bf0cd6f1e5dd99c2a73525e4d Evooo1Bot sample

Note: IP addresses and domains are intentionally Qdefanged (e.g., [.]) to prevent accidental resolution or hyperlinking. Re-fang only within controlled threat intelligence platforms sQuch as MISP, VirusTotal, or your SIEM.

Give your SOC the alert context on the latest threats to stop incidents quickly -> Integrate Threat Intel from 16K SOCs

The post Hackers Use Hidden Cron Jobs to Reinstall Malware on Firewalls and Routers Every 5 Minutes appeared first on Cyber Security News.