The site's security writer, Catalin Cimpanu, has form [1, 2, 3] in screwing up when he writes about Linux. And ZDNet has a person on staff, Stephen J. Vaughan-Nicholls, who knows Linux very well. So why exactly the kind of dross that was published on 24 November was ever allowed to pass the editor's knife is puzzling.
To details. In this case, Cimpanu was writing about a botnet known as Stantinko, a new version of which has apparently been detected by the Israeli security firm Intezer and detailed in a blog post which was shared with Cimpanu before being made available to world+dog.
Before I go any further, let me say that i have reported on Intezer at least thrice, and they are sound when it comes to their research. There is no hyperbole and when they say something, they have enough evidence to do so.
And Cimpanu is surprised that the package is leaner, something that happens quite often when the author of a specific malware package decides to narrow his/her field of operations.
Then, noting that the Intezer folk have pointed out that the Stantinko process on Linux now runs with the name httpd — the same as the world's most widely used Web server software, Apache — Cimpanu claims that Apache is included by default in many Linux distributions. That is not the case; Linux distro installers have software separated into groups and unless one picks a particular group of packages to be installed, it will not be on your machine.
I have been running Debian GNU/Linux for the last 20 years and Apache is not on my workstation because I never chose to install it. I don't need it.
And then, as usual, as in earlier cases, there are a few gems from Cimpanu . One is this: "What Linux server owners need to know is that despite Linux being a secure OS, malware often burrows deep inside systems because of misconfigurations. In Stantinko's case, this botnet goes after server administrators who use weak passwords for their databases and CMSs."
Linux is far from being a secure OS. If that claim is made about NetBSD, FreeBSD or OpenBSD, one would have to agree. And a weak password is not a misconfiguration. It is a stupid mistake. A configuration change is a change in the configuration file for a given package.
Then we have this generalisation: "Malware rarely exploits OS-level vulnerabilities to gain a foothold on a system." This is dead wrong. In a majority of Windows infections, the infiltration occurs because the attackers hone in on system flaws, on components of Windows which are part of the core install and cannot be removed.
Cimpanu goes on: "In most cases, malware gangs usually focus on: app misconfigurations that have left open ports or admin panels exposed online; outdated apps left without security patches; systems/apps that use weak passwords for Internet-facing services; tricking users into taking dangerous actions (social engineering); or exploiting bugs in the apps that run on top of the operating system."
This is, again, highly misleading. Perhaps he wanted to excuse Microsoft for its shoddy programming, but practically all the exploits aimed at Windows target system issues.
There's more. "Exploits in the Linux OS itself are rarely used, and usually after the malware has already gained access to a system through one of the methods above," Cimpanu writes. It is unclear what he is trying to say; when he says Linux OS, what does he mean? Linux refers only to the kernel.
Again, there is a curiously worded paragraph where one has no idea what is being said: "These exploits, employed as second-stage payloads, are usually employed to elevate privileges from low-level to admin accounts, so the malware can take full control of the attacked system. This is why, even if Linux (or other OS) isn't targeted directly, it still needs to run up-to-date versions to prevent these user-to-root elevations once attackers gain a foothold on infected hosts." Does ZDNet employ no sub-editors?
Finally, there is this: "Keeping systems safe from attacks is easy, as most system administrators need to keep apps up-to-date and to use strong passwords. Yet, this is always hard work because, in most cases, companies run hundreds or thousands of systems at the same time, and attackers only need to find one weak link to get in." I doubt that any seasoned sysadmin would claim that keeping a system safe from attacks is easy. And when people run hundreds of servers, they normally write simple scripts to automate most of their daily tasks. But then perhaps Cimpanu is unaware of that. Next time he might like to have a word with his colleague, Vaughan-Nicholls, before sending his copy in.