UPDATE! RB493G - problem with ETH reconnecting

Stunherald · Thu Jun 30, 2011 12:07 am

Hello, i have a problem with my new RB493G and i'm out of my own "guesses", so hope someone help me here.

After first boot, i removed default configuration and set everything as i wanted. Everythig was fine until i plug the uplink cable. I have a WAN port set to Eth2 (because Eth1 is from the side, but all cables going behind the cabinet, so its easier to use another port). The main problem is, the WAN port going UP and DOWN over and over. This is part of the log:

18:24:34 interface,info 2_Uplink link down 18:24:35 interface,info 2_Uplink link up (speed 100M, full duplex) 18:24:36 interface,info 2_Uplink link down 18:24:37 interface,info 2_Uplink link up (speed 100M, full duplex) 18:24:38 interface,info 2_Uplink link down 18:24:42 interface,info 2_Uplink link up (speed 100M, full duplex) 18:24:44 interface,info 2_Uplink link down 18:24:45 interface,info 2_Uplink link up (speed 100M, full duplex) 18:24:48 interface,info 2_Uplink link down 18:24:52 interface,info 2_Uplink link up (speed 100M, full duplex) 18:24:57 interface,info 2_Uplink link down 18:24:59 interface,info 2_Uplink link up (speed 100M, full duplex)

and over and over like this. I thought that i have somewhere some conflict, co i disabled all ports except WAN (Eth2), me (Eth3) and one more station (Eth6). To be sure, i disabled all NAT and FW rules and IGMP proxy. I keep active Bridge, DHCP client for WAN and DHCP server for the bridge. Unfortunately, still the same issue.

LAN connects:

Eth1 - nothing Eth2 - WAN Eth3 - Station (me) Eth4 - Station (Off) Eth5 - Station (Off) Eth6 - Station (ON) Eth7 - nothing Eth8 - WiFi router (temporary, until miniPCI card arrive, also Off during the test) Eth9 - NAS (Off) Ports 3,4,5,6 are in BRIDGE

I was worried about the physical port can be broken, so to be sure, i changed the configuration. I set the Eth1 as WAN port and i added Eth2 to the existing bridge. Then i connect via Eth2. It was a supprise that everything was OK, no port up/down cycles, speed on 1000BaseT full duplex. But then, after few seconds, Eth6 (where another machine was conencted) start "cycling" in "UP/DOWN".

So i start checking every configuration, because that was odd, and i found one possible cause of this problem in SWITCH setting.

# NAME SWITCH VLAN-MODE VLAN-HEADER
0 1_Freeswitch1fallback leave-as-is
1 2_WANswitch2fallback leave-as-is
2 3_aaaswitch2fallback leave-as-is
3 4_bbbswitch2fallback leave-as-is
4 5_cccswitch2fallback leave-as-is
5 6_dddswitch1fallback leave-as-is
6 7_Freeswitch1fallback leave-as-is
7 8_WRouterswitch1fallback leave-as-is
8 9_NASswitch1fallback leave-as-is
9 switch1_cpuswitch1fallback leave-as-is
10 switch2_cpuswitch2fallback leave-as-is

Basicly, port which is on the same switch as WAN port failing, or the wan port failing.

I didn't find any way how to turn off the switch, because i don't need to use this (not now, maybe latter).

So can anyone please help me how to set things up to work properly? (I need keep the bridge set because of some FW/NAT rules.)

谢谢s a LOT for any usefull info

BTW: I can set a guest account, if needed.

//EDIT: CPU overload solved.

TachinOli · Thu Jun 30, 2011 6:49 pm

you tried change the patch cord?

regards.

Stunherald · Thu Jun 30, 2011 8:01 pm

There is no problem with cables. If i put the same cable to another port, it is OK.

Like i said ...
Uplink cable to Eth2 (set as WAN), workstation on Eth3 and Eth6 -> problem on Eth2
Uplink cable to Eth1 (set as WAN), workstation on Eth3 (or Eth2, does not matter) and Eth6 -> problem on Eth6

Cables are OK, but i'm missing something in configuration ...

I'm trying to set up properly config for this, but i'm facing this stupid problem all the time on some port(s)...

1x WAN 4x workstation 1x NAS 1x WiFi router (temporary until miniPCI cart arrive) 1x Server 1x tunel - planned to next month

I need to have that workstations AND the NAS in one subnet. Since bridging is CPU consuming (and still does not working because of the port failing), i tried to set up a switch, but there is not enought port on one switch, because i need 5+1(as master) <-- not possible. Server port plus tunel port will be isolated, about WiFi, it does not matter, its temporary.

I'm rly lost and tired now :/

EDIT:Ok, i finally understodt how those evil switches works ... so my actual configuration is:

Eth1 - master port for Eth6,7,8,9(switch1) Eth5 - WAN Bridge - Eth1,2,3,4 DHCP server on Bridge

EDIT: I'm usingRouterOS 5.5and firmware is2.29

Beccara · Sat Jul 02, 2011 10:03 am

Transferring data between a bridge and non-bridge port does use CPU depending on packet size and amount yes you can see 100% CPU

This is one of the reasons why we moved to using managed switches with RB's hanging off a trunk port since it's impossible to get a RB with more than 5 ports in the same switch group!!

Stunherald · 2011年坐7月2日下午2:12

Ok, thank you, so looks like i'll buy some L2 8port switch and let Mikrotik only do the FW/NAT work + Wireless.

BTW: I still facing the port UP/DOWN "problem" ... the connection on the end-point is not interrupted (like in prev. case), but still ... what's going on? It is HW failture and shoud i replace the RB or it is ROS bug and will be fixed? It does not happening when i had RG450G with ROS 4.9

03:47:29 interface,info 5_WAN link down 03:47:30 igmp-proxy,info stopping IGMP proxy forwarding 03:47:31 interface,info 5_WAN link up (speed 100M, full duplex) 03:47:32 igmp-proxy,info starting IGMP proxy forwarding 04:06:50 interface,info 7_Holy link down 04:06:51 interface,info 1_Master link down 04:06:52 interface,info 7_Holy link up (speed 100M, full duplex) 04:06:53 interface,info 1_Master link up (speed 1000M, full duplex) 04:42:15 interface,info 5_WAN link down 04:42:16 interface,info 5_WAN link up (speed 100M, full duplex) 04:42:16 igmp-proxy,info stopping IGMP proxy forwarding 04:42:16 igmp-proxy,info starting IGMP proxy forwarding 06:26:47 interface,info 5_WAN link down 06:26:48 igmp-proxy,info stopping IGMP proxy forwarding 06:26:49 interface,info 5_WAN link up (speed 100M, full duplex) 06:26:50 igmp-proxy,info starting IGMP proxy forwarding 11:56:47 interface,info 5_WAN link down 11:56:48 igmp-proxy,info stopping IGMP proxy forwarding 11:56:49 interface,info 5_WAN link up (speed 100M, full duplex) 11:56:50 igmp-proxy,info starting IGMP proxy forwarding 12:30:21 interface,info 7_Holy link down 12:30:22 interface,info 1_Master link down 12:30:24 interface,info 7_Holy link up (speed 1000M, full duplex) 12:30:25 interface,info 1_Master link up (speed 1000M, full duplex) (check prev post for information about ports)

Cables are absolutely OK (i checked them all yesterday with Fluke tester).

Stunherald · Mon Mar 05, 2012 8:51 pm

更新我的Mikroti雷竞技网站k无线网卡,所以additional wifi router is not needed anymore, but i'm still having problems with the port falling...

Eth1 - master port for Eth6,7,8,9(switch1)
Eth5 - WAN
Bridge - Eth1,2,3,4,WLAN
DHCP server on Bridge

19:03:55 interface,info 6_test link down
19:03:59 interface,info 6_test link up (speed 1000M, full duplex)
19:07:50 interface,info 6_test link down
19:07:53 interface,info 6_test link up (speed 1000M, full duplex)
19:08:36 interface,info 6_test link down
19:08:40 interface,info 6_test link up (speed 1000M, full duplex)
19:10:29 interface,info 6_test link down
19:10:32 interface,info 6_test link up (speed 1000M, full duplex)

- Cable is OK (on different port - OK, another cable - same problem)
- Station is OK (tried the same workstation on another port - OK, another station on port 6 - same problem)

So, is this HW problem, or FW/SW problem? Or configuration problem / switch limitation ?

FW:2.39
RouterOS:5.14

zeane · Wed Mar 07, 2012 12:50 am

Similar problem with the same RB (493G) Using 5.14, my Ethernet ports set at 1Gbps, connected to various different types of equipment, with Auto Negotiation - Turned on... Then sometimes when rebooting RB, dhcp cannot be passed to device, and winbox says the Ethernet port speed is 100mbps... ?!????

Has anyone seen similar issues? I have reloaded the FW, (latest) etc, and it works ok for a while, but after few reboots, things go a little haywire....

This RB is a vast improvement over the RB450G.... But other issues, as such, are driving me crazy...

Like when I upgraded from 5.10-5.11/ and similarly again when upgrading 5.13-5.14 - was no simple process, it crashed after rebooting (to install upgrades) and I had to get in via serial cable.... Yadda-yadda! ~ Oiii!

I'd be happy to send info to MK or anyone else if further details are required...

NO SOLUTION AS YET!!! Any work-arounds.. Or is this just an issue with v5.14?? And rb493G?

Cheeersssss.....

Anthony@zeane.com

---
Cape Town - South Africa

f4nt0 · Sun Jul 14, 2013 11:55 am

Hello,

I experience the same problem. I've tested RouterOS 5.14 up to 5.25 and 6.0beta up to 6.1 with firmware 3.07 and earlier. WAN port is still going randomly up and down.
Here is the packet capture from today - 4 packets (all MAC & IP addresses were changed):

不。时间源目的地协议长度信息108:16:59.842664000  255.255.255.255 MNDP 203 Source port: rrac Destination port: rrac Frame 1: 203 bytes on wire (1624 bits), 203 bytes captured (1624 bits) on interface 0 Ethernet II, Src: Routerbo_aa:04:37 (00:0c:42:aa:04:37), Dst: AsustekC_b2:27:cd (54:04:a6:b2:27:cd) Internet Protocol Version 4, Src: 192.168.10.1 (192.168.10.1), Dst: 192.168.10.7 (192.168.10.7) User Datagram Protocol, Src Port: 35554 (35554), Dst Port: 37008 (37008) TZSP: Ethernet: Ethernet II, Src: Routerbo_aa:04:33 (00:0c:42:aa:04:33), Dst: Broadcast (ff:ff:ff:ff:ff:ff) Internet Protocol Version 4, Src:  (), Dst: 255.255.255.255 (255.255.255.255) User Datagram Protocol, Src Port: rrac (5678), Dst Port: rrac (5678) Mikrotik Neighbor Discovery Protocol No. Time Source Destination Protocol Length Info 2 08:16:59.842665000 fe80::20c:42ff:feaa:443 ff02::1 MNDP 223 Source port: rrac Destination port: rrac Frame 2: 223 bytes on wire (1784 bits), 223 bytes captured (1784 bits) on interface 0 Ethernet II, Src: Routerbo_aa:04:37 (00:0c:42:aa:04:37), Dst: AsustekC_b2:27:cd (54:04:a6:b2:27:cd) Internet Protocol Version 4, Src: 192.168.10.1 (192.168.10.1), Dst: 192.168.10.7 (192.168.10.7) User Datagram Protocol, Src Port: 35554 (35554), Dst Port: 37008 (37008) TZSP: Ethernet: Ethernet II, Src: Routerbo_aa:04:33 (00:0c:42:aa:04:33), Dst: IPv6mcast_00:00:00:01 (33:33:00:00:00:01) Internet Protocol Version 6, Src: fe80::20c:42ff:feaa:443 (fe80::20c:42ff:feaa:443), Dst: ff02::1 (ff02::1) User Datagram Protocol, Src Port: rrac (5678), Dst Port: rrac (5678) Mikrotik Neighbor Discovery Protocol No. Time Source Destination Protocol Length Info 3 08:16:59.842665000 Routerbo_aa:04:33 CDP/VTP/DTP/PAgP/UDLD CDP 142 Device ID: hq.hq Port ID: wan Frame 3: 142 bytes on wire (1136 bits), 142 bytes captured (1136 bits) on interface 0 Ethernet II, Src: Routerbo_aa:04:37 (00:0c:42:aa:04:37), Dst: AsustekC_b2:27:cd (54:04:a6:b2:27:cd) Internet Protocol Version 4, Src: 192.168.10.1 (192.168.10.1), Dst: 192.168.10.7 (192.168.10.7) User Datagram Protocol, Src Port: 35554 (35554), Dst Port: 37008 (37008) TZSP: Ethernet: IEEE 802.3 Ethernet Logical-Link Control Cisco Discovery Protocol No. Time Source Destination Protocol Length Info 4 08:16:59.842665000 Routerbo_aa:04:33 Broadcast ARP 89 Who has ? Tell  Frame 4: 89 bytes on wire (712 bits), 89 bytes captured (712 bits) on interface 0 Ethernet II, Src: Routerbo_aa:04:37 (00:0c:42:aa:04:37), Dst: AsustekC_b2:27:cd (54:04:a6:b2:27:cd) Internet Protocol Version 4, Src: 192.168.10.1 (192.168.10.1), Dst: 192.168.10.7 (192.168.10.7) User Datagram Protocol, Src Port: 35554 (35554), Dst Port: 37008 (37008) TZSP: Ethernet: Ethernet II, Src: Routerbo_aa:04:33 (00:0c:42:aa:04:33), Dst: Broadcast (ff:ff:ff:ff:ff:ff) Address Resolution Protocol (request)