Bridge mode dhcp lease stopping every ~5-10mins

[firenfoll]

Sorry for the lack of german -  I can try but i'm not sure it's any good for anyone 😄

I have a simple setup with the TG3442DE (MP3 hw version) vodafone kabel modem on the 1000mb/s plan. It is in bridge mode connected to a mikrotik rb5009. It works but I'm essentially getting constant interruptions that seem to be caused by the link going down all the time:

 21:26:34 interface,info ether8-wan link up (speed 1G, full duplex)
 21:26:35 dhcp,info dhcp-client on ether8-wan got IP address 188.194.166.XX
 21:29:00 interface,info ether8-wan link down
 21:29:00 dhcp,info dhcp-client on ether8-wan lost IP address 188.194.166.XX - lease stopped locally
 21:29:04 interface,info ether8-wan link up (speed 1G, full duplex)
 21:29:04 dhcp,info dhcp-client on ether8-wan got IP address 188.194.166.XX
 21:29:22 interface,info ether8-wan link down
 21:29:22 dhcp,info dhcp-client on ether8-wan lost IP address 188.194.166.XX - lease stopped locally
 21:29:24 interface,info ether8-wan link up (speed 1G, full duplex)
 21:29:24 dhcp,info dhcp-client on ether8-wan got IP address 188.194.166.XX
 21:32:31 interface,info ether8-wan link down
 21:32:31 dhcp,info dhcp-client on ether8-wan lost IP address 188.194.166.XX - lease stopped locally
 21:32:37 interface,info ether8-wan link up (speed 1G, full duplex)
 21:32:38 dhcp,info dhcp-client on ether8-wan got IP address 188.194.166.XX
 21:46:44 interface,info ether8-wan link down
 21:46:44 dhcp,info dhcp-client on ether8-wan lost IP address 188.194.166.XX - lease stopped locally
 21:46:50 interface,info ether8-wan link up (speed 1G, full duplex)
 21:46:51 dhcp,info dhcp-client on ether8-wan got IP address 188.194.166.XX

I followed a few ideas:

1. https://forum.vodafone.de/t5/St%C3%B6rungsmeldungen-Internet-TV/Interface-Flapping-nach-umstellung-a... - but the rb5009 has only one switch chip and in any case cpu flow control is disabled and all devices are on 1000BASE-T on my network.

[admin@hydrogen] /interface/ethernet/switch> export verbose 
# nov/01/2021 22:00:04 by RouterOS 7.1rc5
# software id = 9RE1-EXXX
#
# model = RB5009UG+S+
# serial number = XX
/interface ethernet switch
set 0 !cpu-flow-control l3-hw-offloading=no mirror-egress-target=none name=switch1
/interface ethernet switch port
set 0 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 1 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 2 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 3 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 4 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 5 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 6 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 7 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 8 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none
set 9 !egress-rate !ingress-rate mirror-egress=no mirror-ingress=no \
    mirror-ingress-target=none

 2. I tried setting the dhcp client default route selection behaviour to yes and special-classless as defined here: https://wiki.mikrotik.com/wiki/Manual:IP/DHCP_Client

[admin@hydrogen] /interface/ethernet/switch> /ip/dhcp-client/print 
Columns: INTERFACE, USE-PEER-DNS, ADD-DEFAULT-ROUTE, STATUS, ADDRESS
# INTERFACE   USE-PEER-DNS  ADD-DEFAULT-ROUTE  STATUS  ADDRESS          
;;; WAN iface from vodafone modem
0 ether8-wan  yes           special-classless  bound   188.194.166.XX/24

3. Tightened all the cables somewhere a little loose, condition all looks good and installer only did this 4 days ago. I admit that I didn't really test it properly and was just happy it was done so quick.

It looks to me very similar to https://forum.vodafone.de/t5/Archiv-St%C3%B6rungsmeldungen/Verbindung-im-Bridge-Modus-bricht-t%C3%A4... but I'm not getting explicit NAKs. I haven't taken a pcap trace yet but I guess that's the next step.

[Kieferer]

Previously I used the same WAN IP Range with my pfsense on igb0 like you have. As VF HW I have the same Modem with HW Version 1. I do not believe that there is something wrong with th VF HW.
According to your mentioned WAN IP, you might be from Rosenheim and hooked up in Bad Aibling?

Here my Log after a reboot of the VF Station. The HW NIC's on my pfsense are Intel I-211. ->igb

Works since years like a charm.

Mehr anzeigen

Nov 2 10:08:12	php-fpm	630	/rc.linkup: Accept router advertisements on interface igb0
Nov 2 10:08:12	php-fpm	630	/rc.linkup: calling interface_dhcpv6_configure.
Nov 2 10:08:12	php-fpm	630	/rc.linkup: The command '/sbin/dhclient -c /var/etc/dhclient_wan.conf igb0 > /tmp/igb0_output 2> /tmp/igb0_error_output' returned exit code '11', the output was ''
Nov 2 10:08:12	kernel		pid 41803 (dhclient), jid 0, uid 65: exited on signal 11
Nov 2 10:08:09	kernel		arpresolve: can't allocate llinfo for 178.26.168.xxx on igb0
Nov 2 10:08:07	kernel		arpresolve: can't allocate llinfo for 178.26.168.xxx on igb0
Nov 2 10:08:06	check_reload_status	668	Linkup starting igb0
Nov 2 10:08:06	kernel		igb0: link state changed to UP

What kind of NICs are used in your HW?

Of course you have read this as well

[firenfoll]

I'm using the RB5009 directly connected to the modem which is just a simple ipv4 NAT setup right now. As mentioned I checked the RB3011 post but dont think that applies since I have only one switch chip, everything is on 1000base-t and cpu offloading is turned off. On the network I have a netgear GS108, a mikrotik cap AC, a RB2011 and an i350 NIC, all of which where on my previous network with 1&1 with no issue, the RB5009 is new though, but I switched it with the RB2011 that I used for years (too slow to route ~1GB traffic) and have similar issues.

I'm in the Ulm area so other direction than Rosenheim. It was working horribly this morning and now for the last 5hrs it has been working perfectly 😕 So it seems a little like the doscis connection is tempermental?

[Kieferer]

mhmm.

I gues your issue is more to microtik relatet.

All my used VF Modems - Hiltron, Compal and a various Series of VF Stations - draw flawless my WAN adresses. I'm using unbound and I'm also in Bridge Mode.

Here a example from my pfsense. Last reboot done at 8:00. Every 2700s this:

Mehr anzeigen

Nov 2 13:34:57	dhclient	86492	bound to 178.26.168.xxx -- renewal in 2700 seconds.
Nov 2 13:34:57	dhclient	8219	Creating resolv.conf
Nov 2 13:34:57	dhclient	7865	RENEW
Nov 2 13:34:57	dhclient	86492	DHCPACK from 83.169.184.130
Nov 2 13:34:57	dhclient	86492	DHCPREQUEST on igb0 to 83.169.184.130 port 67

Can you find someting similar in your logs?

67 – UDP Bootstrap Protocol (BOOTP) Server; auch genutzt von DHCP

The only thing what I have added in my loader.conf was intel relatet stuff, more cosmetic nature

Mehr anzeigen

kern.cam.boot_delay=10000
kern.ipc.nmbclusters="1000000"
kern.ipc.nmbjumbop="524288"
kern.ipc.nmbjumbo9="524288"
legal.intel_iwi.license_ack=1
legal.intel_ipw.license_ack=1
opensolaris_load="YES"
zfs_load="YES"
kern.geom.label.disk_ident.enable="0"
kern.geom.label.gptid.enable="0"
opensolaris_load="YES"
zfs_load="YES"
kern.ipc.semopm=100
kern.ipc.semmni=128
kern.ipc.semmns=32000
kern.ipc.shmmni=4096
boot_multicons="YES"
boot_serial="YES"
console="comconsole,vidconsole"
comconsole_speed="115200"
autoboot_delay="3"
net.link.ifqmaxlen="128"

But I gues, this won't solve your issue.

Also I have read that Fritzbox users are also complaining.

Did find such VF DHCP WAN issues in the microtik forum?

[firenfoll]

I took a pcap trace of the dhcp when the interface goes down but seems all normal.

0.000000 0.0.0.0 255.255.255.255 DHCP 342 DHCP Discover - Transaction ID 0x5fb96bfd
0.028581 88.134.229.xxx 255.255.255.255 DHCP 342 DHCP Offer - Transaction ID 0x5fb96bfd
0.028753 0.0.0.0 255.255.255.255 DHCP 342 DHCP Request - Transaction ID 0x5fb96A
0.050977 88.134.229.130 255.255.255.255 DHCP 342 DHCP ACK - Transaction ID 0x5fb96bfd.

I changed the cables, the modem port etc, all also seems good. Guess I can also change the mikrotik port, but i'm a bit out of ideas at this point.

[firenfoll]

Out of desperation I put a cheap gigabit switch between the two devices and seems its now working fine. I have really no idea why a random switch in between would help but somehow it does.

It does seem like more of a mirkotik port flapping issue than the vodafone modem, so I opened a support ticket with them - let's see 😕

[Kieferer]

Which DHCP server is active?

Microtik or the switch. Some switches have a build in DHCP server. I.e Netgear GS108eV3.

[firenfoll]

The DHCP client is on the RB5009 ether8 port, the switch is an asus GX-D1051 that I had lying around, it's for sure not a managed unit.