On Fri, 2016-10-07 at 17:20 +0200, Jann Haber wrote: > It sounds exactly like the bug described here for Kernel 4.1.28: > http://lkml.iu.edu/hypermail/linux/kernel/1607.1/05446.html > > Is it possible, that this bug has been backported to the current > Debian > stable kernel? The Debian stable kernel, along with all the upstream stable branches, has added several fixes to validation of iptables and ip6tables rules, and one of these is known to cause a performance regression for systems using large numbers of rules. I aim to fix this in the next stable or security update, but I don't have a timetable for that. Ben. -- Ben Hutchings Every program is either trivial or else contains at least one bug
Attachment:
signature.asc
Description: This is a digitally signed message part