What's new

Syslog Flooded With nf_conntrack expectation table full Errors, RT-AC86U

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Prodeje79

Regular Contributor
I think I may have posted my thread in the incorrect area yesterday!
I just unboxed my brand new replacement RT-AC86U from ASUS after my original died in 3 months.

This time I was extra careful during the setup. Followed nuclear reset twice. Then did m&m config.
Device came with 3.0.0.4.384_32799 and I put on merlin 384.19
This then spams the logs:
Jan 11 16:18:54 kernel: nf_conntrack: expectation table full
over and over.....

I've done nothing else and am seeing this in the logs already like my prior device. Ultimately once I installed FlexQoS, it would clear it, but then in about 20-30 mins it is full again.

I had followed the solved thread previously, but I don't think we got an official response from Merlin himself.
thread was: https://www.snbforums.com/threads/solved-syslog-flooded-with-nf_conntrack-errors-rt-ax88u.62501/

Love to get the scoop and if this can be fixed right from the flash without having to do anything.
Else, what for sure should I do now?

One fix was to tweak a number to 240 or 384..... "Value 240 written to nf_conntrack_expect_max"

I've seen others posting their logs with this in it to. Are we just ignoring or is this something that @RMerlin can fix?

Thanks!
 
Not a firmware issue per se, so no need for RMerlin to respond, and only noticed by a handful of users (ie doesnt affect everyone).

3 options to 'fix'

1) connect your clients 1 by 1 and figure out which one of your clients is causing the conntracking to fill up and deal with it instead of adjusting router conntrack settings to accommodate
2) set a higher value to nf_conntrack_expect_max (default is 54ish on the 86U)
3) just ignore them if you arent noticing connectivity issues
 
Thank you! I have a few IoT things, tons of google stuff etc, Synology NAS, but nothing unusual. Are there any risks or considerations raising this? what are the possible values etc? Ill try the 240 though, just curious....

What about these modprobe modules.dep items that pop up in the log after?
Just ignore?
Jan 12 19:59:03 conntrackscript: Value 240 written to nf_conntrack_expect_max, restarting conntrack
Jan 12 19:59:05 rc_service: service 24698:notify_rc restart_conntrack
Jan 12 19:59:05 modprobe: module nf_conntrack_proto_gre not found in modules.dep
Jan 12 19:59:05 modprobe: module nf_nat_proto_gre not found in modules.dep
Jan 12 19:59:05 modprobe: module nf_conntrack_pptp not found in modules.dep
Jan 12 19:59:05 modprobe: module nf_nat_pptp not found in modules.dep
Jan 12 19:59:05 custom_script: Running /jffs/scripts/service-event-end (args: restart conntrack)
 
Cant find the post at the moment, but RMerlin mentioned the module messages can be ignored, its part of the kernel now I believe and the logging is oversight by Asus

I ran a handful of values between default and 1028(experimented) and never noticed any issue, but saying that... to be safe and not cause issues with other conntrack tracking potentially you'd be better off incrementally increasing it till you're satisfied
 
Last edited:

Similar threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top