4 ms·
Sshguard protects networked hosts from brute force attacks
- simple10 11y agoAnybody know if this is better than fail2ban?
- grhmc 11y agoBased on the feature set it looks like a me-too of fail2ban. I'd be interested in seeing a comparison. The site's FAQ has no mention of fail2ban.
- alex_duf 11y agoAfter a 5s read of the page, it looks like fail2ban, but with less features...
- fasteo 11y ago>> sshguard is different from the other two in that it is written in C, is lighter and >> simpler to use with fewer features while performing its core function equally well.[1] [1] https://wiki.archlinux.org/index.php/Sshguard https://wiki.archlinux.org/index.php/Sshguard
- bradleyland 11y agoI wouldn't say better. More of a lighter weight alternative. Fail2ban can do lots of stuff. You can define fail2ban rules for any service that writes login failures to a log file. Sshguard, on the other hand, endeavors only to monitor ssh. This makes it more focused, and therefore smaller. Whether this is "better", depends on what you need. I don't know about your systems, but I use fail2ban for protecting several different administrative systems, so I'd still need it. Thus, I won't be switching. However, if you only need SSH monitoring, this is worth a close look.
- onestone 11y agosshguard can protect IPv6 as well.
- MichaelGG 11y agoSo does disabling password auth and using real keys. My only experience with sshguard is via Google Cloud Platform. For some reason they enable it by default. Thus it locked me out somehow, making me spend an extra few hours wondering why SSH stopped working. If you're that upset about auth logs, put SSH on another port.
- nieve 11y agoThe big win for me is that it cuts down log noise from ssh brute force attempts. Used with pf/netfilter you get even less. I'm not sure of getting FreeBSD nightly security reports that freeze a mail client for a few seconds as they render and sshguard cuts that down to a manageable block message.
- MichaelGG 11y agoSo don't log those things or especially don't email them? They're obviously not providing useful or actionable information. Just seems like the hassle of getting locked out far outweighs the reduction in a log you're not really monitoring anyways. And doesn't moving ports also kill that junk traffic?