I just started setting up my new RB4011iGS+5HacQ2HnD. I have a Cisco GigE SFP as the WAN port and ether1 as my primary LAN connection. I have setup some basic firewall rules, and my IPv6 tunnel but nothing else. Download speeds should be around 300 Mbps, but I getting much lower than I was when using the Cisco RV340W I am replacing.
Any advice you can offer would be most appreciated.
I did indeed, fresh out of the box configuration with only the masquerade rules, and I get the same speed. I decided to fall back from the SFP , tried a single interface (ether1), then tried a bond of two ports (9/10) and all get about the same performance.
Some questions:
I will assume that your WAN connection is Fiber from Virgin Media – so is the Cisco GigE SFP module white-listed [approved] by Virgin?
Do you know the SYNC rate that Virgin is using … is it 1.25G or 2,5G?
From your description of the issue my assumption would point to the wrong transceiver sync rate which usually means you have to use a module that is approved by Virgin for it work properly. The Other issue is that The RB4011 does not support Passive DAC modules and SFP GPON modules — so this usually means that if you want to use the RB4011 SFP+ port for WAN you cannot do that — instead you will need to use one of your Ethernet Ports like ether1 … and the connectivity chain for that method is via a Media Converter that connects to Virgin using an approved SFP module and from the Media Converter to ether1 on the RB4011.
I have no experience with Virgin Media – so I am just giving you some possibilities. Perhaps some other user who does have experience using Virgin and the RB4011 can chime in with how they establish connectivity.
This is a simple Copper Cisco GLC-TE 1000BASE-T interface. I am using it purely for port-availability on the rest of the ether interfaces.
Additionally, I have tried with ether1, as well as a bond (just to see if the rate would double), however the rate of throughput is nearly constant (under 10Mbps).
I ended having to swap the cisco device back in. Having tried every permutation of interfaces, MSS clamping, doing packet captures (shed loads of DUP ACKs)
No avoidance was intended. I thought I had answered by stating that this is not an optical connection. In addition, I eliminated the SFP from the issue by reverting to utilising ether1 as you suggested.
There is no SFP on the Virgin Media side of things.
If you are curious about the ISP connection parameters, here is what I have from their DOCIS3.0 modem:
General Configuration
Network access Allowed
Maximum Number of CPEs 10
Baseline Privacy Enabled
DOCSIS Mode DOCSIS 3.0
Config file CH7465LG_cm_res005_v4-sip.bin
Primary Downstream Service Flow
SFID 200126981
Max Traffic Rate 10485760 bps
Max Traffic Burst 16320 bytes
Min Traffic Rate 0 bps
SFID 200126980
Max Traffic Rate 160000 bps
Max Traffic Burst 3044 bytes
Min Traffic Rate 12000 bps
SFID 200126979
Max Traffic Rate 385200000 bps
Max Traffic Burst 42600 bytes
Min Traffic Rate 0 bps
Primary Upstream Service Flow
SFID 200118791
Max Traffic Rate 2097152 bps
Max Traffic Burst 16320 bytes
Min Traffic Rate 0 bps
Max Concatenated Burst 16320 bytes
Scheduling Type Best Effort
SFID 200118790
Max Traffic Rate 160000 bps
Max Traffic Burst 3044 bytes
Min Traffic Rate 12000 bps
Max Concatenated Burst 1522 bytes
Scheduling Type Best Effort
SFID 200118786
Max Traffic Rate 38520000 bps
Max Traffic Burst 42600 bytes
Min Traffic Rate 0 bps
Max Concatenated Burst 42600 bytes
Scheduling Type Best Effort
Is you Cable Gateway in Bridge mode?
The power levels are excellent and your RoS code looks good to me. Confirm that your Ethernet cable is good.very strange that you’re not getting the throughput … perhaps a defective port, try another port.
Also make sure to shutdown your Cable modem, wait for 5 minutes then repower … that may force a new profile for a different device especially when bridge mode is used.
I am indeed in bridge mode. And as the MAC of the device to which DHCP has assigned an address from the upstream device is cached on the modem, every device or port change necessitates a power down of the Cable modem in order for address assignment to work. The connectivity permutations I have tried so far are as follows:
Copper SFP as WAN link
Ether1 as WAN link
bond of ports ether5 and ether10 as WAN link
Each of these configurations, with no other parameters than a WAN list masquerade resulted in proper throughput.
I did achieve nearly 30 Mbps when I added MSS PMTU clamping to the firewall mangle table.
I also performed a packet capture whilst running a speedtest (speedtest.net to an ISP within 5 km of me) and saw massive numbers of DUP ACKs and retransmission.
Something is definitely wrong with your RB4011 … try one more process of running netinstall-6.45.1 … this will eliminate any form of corruption in the NAND memory and reinstall ROS – then proceed with your Copper SFP as WAN link to see if that works – it should and if it does not then I would suggest that you send MikroTik support your supout.rif file for their observations and suggestion – if this is a new unit or under warranty you may be able to RMA it based on MikroTik Support’s observations of the supout.rif.
So holy shit on shingle it was all on the ISP. After I swapped things back, my speeds were slow. I opened a ticket with them, and have a scheduled appointment with them on Thursday. “Magically” the speed improved a few minutes ago on the old router. I swapped the mikrotik back in, and the speeds were as follows:
PING ms
5
DOWNLOAD Mbps
199.07
UPLOAD Mbps
23.25
So the whole freaking time, it was the ISPs fault!!!