[Bug 1852072] [NEW] Conntrackd segfault when flushing tables

[Ralf Spenneberg]

Public bug reported:

Hi,
I have a setup using conntrackd and keepalived for a ha-firewall. When stress-testing the system by stopping and starting keepalived eventually the conntrackd segfaults.

Nov  4 15:29:46 host conntrackd[1122]: [Mon Nov  4 15:29:46 2019] (pid=1122) [notice] committing all external caches
Nov  4 15:29:46 host Keepalived_logscript: user.info GROUP VG11 MASTER 0
Nov  4 15:29:46 host kernel: [20484.383544] show_signal_msg: 7 callbacks suppressed
Nov  4 15:29:46 host kernel: [20484.383546] conntrackd[1122]: segfault at 10 ip 00005649a5693ab4 sp 00007ffeb5aafec0 error 4 in conntrackd[5649a5688000+34000]

It does not segfault every time. Sometimes it takes 5-10 iterations
starting and stopping the keepalived daemon which will invoke conntrackd
to flush the caches and subsequently segfault.

Kernel: 4.15.0-55-generic #60-Ubuntu SMP Tue Jul 2 18:22:20 UTC 2019
x86_64 x86_64 x86_64 GNU/Linux

Conntrackd: 1.4.4+snapshot20161117

** Affects: conntrack-tools (Ubuntu)
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to conntrack-tools in Ubuntu.
https://bugs.launchpad.net/bugs/1852072

Title:
  Conntrackd segfault when flushing tables

Status in conntrack-tools package in Ubuntu:
  New

Bug description:
  Hi,
  I have a setup using conntrackd and keepalived for a ha-firewall. When stress-testing the system by stopping and starting keepalived eventually the conntrackd segfaults.

  Nov  4 15:29:46 host conntrackd[1122]: [Mon Nov  4 15:29:46 2019] (pid=1122) [notice] committing all external caches
  Nov  4 15:29:46 host Keepalived_logscript: user.info GROUP VG11 MASTER 0
  Nov  4 15:29:46 host kernel: [20484.383544] show_signal_msg: 7 callbacks suppressed
  Nov  4 15:29:46 host kernel: [20484.383546] conntrackd[1122]: segfault at 10 ip 00005649a5693ab4 sp 00007ffeb5aafec0 error 4 in conntrackd[5649a5688000+34000]

  It does not segfault every time. Sometimes it takes 5-10 iterations
  starting and stopping the keepalived daemon which will invoke
  conntrackd to flush the caches and subsequently segfault.

  Kernel: 4.15.0-55-generic #60-Ubuntu SMP Tue Jul 2 18:22:20 UTC 2019
  x86_64 x86_64 x86_64 GNU/Linux

  Conntrackd: 1.4.4+snapshot20161117

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/conntrack-tools/+bug/1852072/+subscriptions