New PPPoE backend, some feedback
-
Yes, I have enabled system/advanced /networking marked Use if_pppoe kernel module for PPPoE client.It seems like a bug in the new version 2.8
-
Yes, there is a bug there. However currently the bug is that the mpd5 specific options are not hidden correctly. And that the cron-job is still set even thought the script it calls is not present when if_pppoe is enabled.
If you need this for now you should add your own cronjob. This should work:
/etc/rc.interfaces_wan_configure opt2
Where opt2 is the pppoe interface internal name.This should probably also be a feature request. Let me see here...
-
The WAN interface is pppoe1 which is on igb1.
There is already a cron job with the script /var/etc/pppoe_restart_pppoe1.
If you can tell me if this is ok or if this script needs to be modified?
-
That script is created to bring up mpd5 so it's not valid when using if_pppoe.
Here's a bug the covers the invalid options: https://redmine.pfsense.org/issues/16155
I'll open a feature request to add back the periodic reset for if_pppoe.
-
Thanks for reporting.
I hope this feature will be included in pfSense 2.8. -
It probably won't be in 2.8 but it's a pretty simple feature. I'd expect it could be added via system patches after release.
Or you can still just add the cronjob manually.
-
The thing is I don't know how to manually make the script to put in cron.
If you can kindly give me a working example that would be great and would solve the problem permanently.
Thanks -
Edit the cronjob and change the command to:
/etc/rc.interfaces_wan_configure opt2Where opt2 is the internal interface name for your pppoe interface. If that's actually WAN then use
wan. -
@stephenw10 said in New PPPoE backend, some feedback:
/etc/rc.interfaces_wan_configure
I'll ask something else while we're at it...
Regarding manually starting and stopping the WAN interface, which is PPPoE — what's the proper way to do it? Especially with the new backend...
/etc/rc.linkup stop wan
/etc/rc.linkup start wanOr is there something newer now?
-
It worked fine.
Thanks for your help. -
That should work. There's nothing newer I'm aware of. Because it's now a regular interface (mostly) you should be able to do regular interface things with it. If you want/need to.
-
Is there a definitive list of PPPoE performance tweaks that should be removed before using if_pppoe?
If so, it may be worth adding it to Jim's article too.
Apologies if it exists somewhere but I did look... and clearly I have forgotten all the tweaks I did apply to try and get the best PPPoE performance on the current backend.
️ -
Does anyone have packet loss on a new backend? The connection is not interrupted just loss of packets, about 5%, once a day or so.
MONITOR: WAN_PPPOE has packet lossThis could be something on isp or on my side also, because I haven't seen this before. IDK...
-
Not seen any significant loss here. Been solid since I switched both my edge PPPoE links to it:
And I'd been running it for months before that on one WAN in 25.03.
-
@RobbieTT said in New PPPoE backend, some feedback:
Is there a definitive list of PPPoE performance tweaks that should be removed before using if_pppoe?
There isn't. The testing we did though was without any of the previous pppoe tweaks AFAIK. So the default sysctls only.
-
@stephenw10 said in New PPPoE backend, some feedback:
@RobbieTT said in New PPPoE backend, some feedback:
Is there a definitive list of PPPoE performance tweaks that should be removed before using if_pppoe?
There isn't. The testing we did though was without any of the previous pppoe tweaks AFAIK. So the default sysctls only.
I know, I just need a nudge on what I now need to remove (eg deleting the top entry for a start):
️ -
I think only the net.isr.dispatch value there is non-default.
Default values I see here are:
Tunable Name Description Value net.inet.ip.portrange.first 1024 net.inet.tcp.blackhole Do not send RST on segments to closed ports 2 net.inet.udp.blackhole Do not send port unreachables for refused connects 1 net.inet.ip.random_id Assign random ip_id values 1 net.inet.tcp.drop_synfin Drop TCP packets with SYN+FIN set 1 net.inet.ip.redirect Enable sending IP redirects 1 net.inet6.ip6.redirect Send ICMPv6 redirects for unforwardable IPv6 packets 1 net.inet6.ip6.use_tempaddr Create RFC3041 temporary addresses for autoconfigured addresses 0 net.inet6.ip6.prefer_tempaddr Prefer RFC3041 temporary addresses in source address selection 0 net.inet.tcp.syncookies Use TCP SYN cookies if the syncache overflows 1 net.inet.tcp.recvspace Initial receive socket buffer size 65228 net.inet.tcp.sendspace Initial send socket buffer size 65228 net.inet.tcp.delayed_ack Delay ACK to try and piggyback it onto a data packet 0 net.inet.udp.maxdgram Maximum outgoing UDP datagram size 57344 net.link.bridge.pfil_onlyip Only pass IP packets when pfil is enabled 0 net.link.bridge.pfil_member Packet filter on the member interface 1 net.link.bridge.pfil_bridge Packet filter on the bridge interface 0 net.link.tap.user_open Enable legacy devfs interface creation for all users 1 net.link.vlan.mtag_pcp Retain VLAN PCP information as packets are passed up the stack 1 kern.randompid Random PID modulus. Special values: 0: disable, 1: choose random value 347 net.inet.ip.intr_queue_maxlen Maximum size of the IP input queue 1000 hw.syscons.kbd_reboot enable keyboard reboot 0 net.inet.tcp.log_debug Log errors caused by incoming TCP segments 0 net.inet.tcp.tso Enable TCP Segmentation Offload 1 net.inet.icmp.icmplim Maximum number of ICMP responses per second 0 vfs.read_max Cluster read-ahead max block count 32 kern.ipc.maxsockbuf Maximum socket buffer size 4262144 net.inet.ip.process_options Enable IP options processing ([LS]SRR, RR, TS) 0 kern.random.harvest.mask Entropy harvesting mask 351 net.route.netisr_maxqlen maximum routing socket dispatch queue length 1024 net.inet.udp.checksum compute udp checksum 1 net.inet.icmp.reply_from_interface ICMP reply from incoming interface for non-local packets 1 net.inet6.ip6.rfc6204w3 Accept the default router list from ICMPv6 RA messages even when packet forwarding is enabled 1 net.key.preferred_oldsa 0 net.inet.carp.senderr_demotion_factor Send error demotion factor adjustment 0 net.pfsync.carp_demotion_factor pfsync's CARP demotion factor adjustment 0 net.raw.recvspace 65536 net.raw.sendspace 65536 net.inet.raw.recvspace Maximum space for incoming raw IP datagrams 131072 net.inet.raw.maxdgram Maximum outgoing raw IP datagram size 131072 kern.corefile Process corefile name format string /root/%N.core kern.crypto.iimb.enable_aescbc 1 kern.crypto.iimb.enable_multiq 1 kern.crypto.iimb.use_task 0 kern.crypto.iimb.arch auto kern.crypto.iimb.prefetch 1 kern.crypto.iimb.max_jobs 256Urgh, formatting fail!
-
Well I'm up and running on if_pppoe and I can see the pppoe load being spread across multiple cores. It works!
UK / Openreach FTTP / 1500MTU (baby-jumbos) / Xeon D-1736NT CPU @ 2.70GHz / HyperThreading Off / IPV4 & IPV6 / bidirectional fq_codel / 10 GbE LANs & VLANs
I think I will need to get the Netgate 6100 out and try this. Running pfSense+ on my Xeon provided no issues for single-core PPPoE anyway but it does seem to run at a slightly lower CPU load with the latest config.
No PPP logs or entries in the System logs, which I think has been mentioned already.
️ -
Ooo you got full size packets over pppoe working. For some reason I can't seem to set that. I'm wondering if I'm hitting some hardware restriction.
Did you have to do anything special?
-
@stephenw10 said in New PPPoE backend, some feedback:
Ooo you got full size packets over pppoe working. For some reason I can't seem to set that. I'm wondering if I'm hitting some hardware restriction.
Did you have to do anything special?
Didn't change anything as I did an in-place update so my config was unchanged from before:
@Smaug ~ % ping -D -s 1472 8.8.8.8 PING 8.8.8.8 (8.8.8.8): 1472 data bytes 1480 bytes from 8.8.8.8: icmp_seq=0 ttl=119 time=8.628 ms 1480 bytes from 8.8.8.8: icmp_seq=1 ttl=119 time=8.637 ms 1480 bytes from 8.8.8.8: icmp_seq=2 ttl=119 time=8.457 ms 1480 bytes from 8.8.8.8: icmp_seq=3 ttl=119 time=8.303 ms 1480 bytes from 8.8.8.8: icmp_seq=4 ttl=119 time=8.383 ms 1480 bytes from 8.8.8.8: icmp_seq=5 ttl=119 time=8.600 ms 1480 bytes from 8.8.8.8: icmp_seq=6 ttl=119 time=8.442 ms 1480 bytes from 8.8.8.8: icmp_seq=7 ttl=119 time=8.582 ms 1480 bytes from 8.8.8.8: icmp_seq=8 ttl=119 time=8.845 ms 1480 bytes from 8.8.8.8: icmp_seq=9 ttl=119 time=8.423 msSo everything worked for me out of the gate; I only checked it 'just to be sure' as it is a bit of a UK oddity.
I can compare a few things for you tomorrow if you like.
️
