1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.


Discussion in 'Cisco Small Business Routers and VPN Solutions' started by mhakman, Oct 6, 2005.

  1. mhakman

    mhakman Network Guru Member

    I’m looking for information about WRV54G, RV082, RV042 and also BEFSX41 and BEFVP41.

    1. Performance data for LAN-LAN, LAN-WAN and WLESS with and without VPN. Both throughput and delays are interesting.
    2. Do any of the routers support VPN using AES? How many bits?
    3. Do wireless routers support WPA-PSK AES?
    4. Do any of the routers support so called port restricted cone NAT?
    5. Do they support so called consistent NAT? Results from the natcheck program run would be most interesting.
    6. Do all routers support UPnP?
    7. What are the differences between RV082 and RV042 apart from the obvious 4/8 port switch?
    8. What are the differences between (W)RV and BEF series?

    Any other information that could be useful in making a purchase decision such as stability, reboots, support etc would also be gratefully appreciated.

  2. net_eng

    net_eng Network Guru Member

    I can only answer for RV042 and these are quick answers sorry, if you want more details just let me know.

    1. Performance(firewall enabled)



    2. RV042 supports DES, 3DES, AES-128, AES-192, AES-256 for VPN
    3. Not wireless
    4. Will get back to you

    5. Natcheck
    TCP consistent translation: YES (GOOD for peer-to-peer)
    TCP simultaneous open: YES (GOOD for peer-to-peer)
    TCP loopback translation: NO (BAD for P2P over Twice-NAT)
    TCP unsolicited connections filtered: YES (GOOD for security)

    UDP consistent translation: YES (GOOD for peer-to-peer)
    UDP loopback translation: NO (BAD for P2P over Twice-NAT)
    UDP unsolicited messages filtered: YES (GOOD for security)

    6. Supports Upnp
    7. Processor and Memory I do believe, RV082 allows you to manage the switch ports individually(change duplex, speed etc) RV042 does not as of firmware
    8. Can't answer for wireless

    In terms of stability, reboots etc. I have used my RV042 for the past 8 months and have never had it reboot, crash or any other problem at all. Thats my experience though I have seen others have many problems.

    Its really bad luck if you buy a router(any) and it crashs, reboots, and you have many problems. Almost like you bought a lemon run router. Otherwise every single person would have problems and I can tell you I have had none in the past 8 months of use.

    Just my 2 cents, hope this helps.
  3. net_eng

    net_eng Network Guru Member

    Sorry in my answer to question 1 I forgot to include non vpn and vpn

    For NON VPN:



    For VPN
    I have a VPN from one lab to another on the same site so it has maximum available bandwidth.

    VPN-to-VPN(gateway to gateway) on the WAN side

    Throughput 2MB/s(thats megabytes per second)
  4. DocLarge

    DocLarge Super Moderator Staff Member Member

    Which particular tools are you currently using to get your stats with?

  5. mhakman

    mhakman Network Guru Member

    Many thanks net_eng!

    From the way in which you specify throughput (megabytes/sec) I assume that you are measuring it by transferring a file and taking the time it took. Is this a correct assumption? If so, how big is the file? 9.8 MB/sec is about 78.4 Mbps. Adding file transfer overhead to this and assuming your LAN speed of 100 Mbps indicates that the router in non-VPN mode is about as fast as the LAN. This sounds quite impressive.

    In VPN mode your numbers give 16 Mbps plus file transfer overhead. This is very good too.

    I found a review of RV082 at www.tomsnetworking.com. This review gives about 81.6 Mbps for LAN/LAN, 20 Mbps for non-VPN LAN/WAN, and 20 Mbps for VPN LAN/WAN. Except for the non-VPN LAN/WAN these numbers are consistent with yours given that RV082 should be faster than RV042. There must be however some error in non-VPN LAN/WAN numbers either in yours or in Toms measurement.

    From the natcheck results (question #5) it seems that the router uses port restricted cone NAT (question #4). I’m assuming here that you were using NAT when running natcheck. Could it be that the router internally uses an implementation of the famous netfilter/iptables software?
  6. net_eng

    net_eng Network Guru Member

    I used a 50 MB file and transfered with ftp(from within RV042 network, I also did the same when transferring to my machine through the RV042)

    I am using the RV042 with NAT(private IP internal, public external). The throughput is what ftp gives back to me after the transfers. Now ftp usually is faster than other transfer methods(its supposed to be) but it does give you an idea of what speed you can get if the network is fairly clean and clear of congestion. My numbers for non-vpn and vpn are both with ftp transfers and a clean network. Through the internet I assume it would be much worse(its not as clean as a local network).

    The RV series routers use jungo software(www.jungo.com) and the firewall is based on IPF or IP filter not iptables.

  7. net_eng

    net_eng Network Guru Member

    I went to see tomsnetworking from the URL you gave and I cant figure out how he only got 20Mbps for LAN-WAN. If the vpn performance is 20Mbps how the heck could the LAN-WAN non vpn be the same or slower?

    It may be older firmware but I can tell you my LAN-WAN numbers above are what FTP give back to me when I transfer a 50MB file to an ftp server. It transfers in seconds. I think the programs he used try to simulate normal internet traffic, file sharing etc I dont know.
  8. mhakman

    mhakman Network Guru Member

    To measure throughput Tom sends 0.1 MB file over and over again under 10 seconds. His 20 Mbps measured throughput yields in 10 * 20 / 8 / 0.1 = 250 file transfers during these 10 seconds. Assuming that the router’s transfer speed is 80 Mbps gives transfer time of 0.1 / (80 / 8 ) = 0.01 seconds for 1 file transfer. The total transfer time of 10 seconds is then given by 250 * (0.01 + ts) where ts is the time to setup one file transfer. Solving the equation 10 = 250 * (0.01 + ts) gives ts = 0.03 seconds. Could the time to setup one file transfer be 30 milliseconds? I think so. This time consists of both the setup of file transfer at both ends and of setting up a firewall session within the router itself so 30 milliseconds doesn’t sound too much at all.

    On the other hand if you transfer one 50 MB file then that one 0.03 second overhead wouldn’t be noticeable and you would end with 80 Mbps measured throughput.

    Could that be an explanation of the difference between Toms and your numbers?
  9. mhakman

    mhakman Network Guru Member

    IPF is fine too :D
  10. mhakman

    mhakman Network Guru Member

    Are you sure RV042 supports AES? I'm evaluating RV082 right now and there is no AES there.


Share This Page