You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello,
this is not really an installation issue per se but more like a question related about the synergy between tools such as fail2ban and crowdsec.
My understanding was that nginx ultimate bad bot blocker also contains a list of IPs which are somewhat up-to-date, is that the case or the main intention of the tool is to reject attacks mainly based on patterns/ua and leave the bad IPs as a secondary feature?
Are then fail2ban/crowdsec overkill against nginx logs?
The text was updated successfully, but these errors were encountered:
@maxdd - I always used UBBB with fail2ban (which was already detecting & blocking clients who generated too many bad 40x's)
I think the proper place for rate limiting requests is in the firewall (all web clients should behave nicely)
I prefer nftables firewalls as it's a single config for ipv4 / ipv6 & ipset functionality is built in.
Blocking bad ip's in ipsets at the nftables ingress level hook is a good choice for performance (i.e before packets traverse connection tracking & before prerouting) - requires kernel 5.10+ & nftables 0.9.7+
CSF firewall uses iptables - I've never used CSF so cannot really comment. It does look like a good choice if you don't know how to drop common attacks / like to have a GUI.
plain nftables with fail2ban will probably always be a bit faster (as it's tuned to your environment)
I block 5-6000 persistent ip's with no issues with nftables & fail2ban:
Reading journald directly in fail2ban will also always be quicker
I always intended UBBB to have blocking done by ipset / fail2ban (& not by nginx) - as it's much, much faster (& you can drop bad packets at the ingress level in nftables before they hit conntrack preserving server resources)
I always only had nginx log the 405s for fail2ban to pick up & ban.
Hello,
this is not really an installation issue per se but more like a question related about the synergy between tools such as fail2ban and crowdsec.
My understanding was that nginx ultimate bad bot blocker also contains a list of IPs which are somewhat up-to-date, is that the case or the main intention of the tool is to reject attacks mainly based on patterns/ua and leave the bad IPs as a secondary feature?
Are then fail2ban/crowdsec overkill against nginx logs?
The text was updated successfully, but these errors were encountered: