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

Cannot get performance with different MAC series (Cisco, 68-7f-74, 00-25-9C) in WRT160N with WDS

Discussion in 'Tomato Firmware' started by xdmv, Aug 14, 2012.

  1. xdmv

    xdmv LI Guru Member

    Last floor is driving me crazy...
    Any help will be very apppreciated :)


    - Cisco E3000, 98:FC:11:95:54:3F (Floor 07), linked with WRT160N (Floor 06), 00:25:9C:68:29:FF
    Rate from E3000 (07) to WRT160N (06): 162~216 Mbps (aprox.)
    Rate from WRT160N (06) to E3000 (07): 216~162 Mbps (aprox.)

    - WRT160N, 00:25:9C:68:29:FF (Floor 06), linked with WRT160N (Floor 05), 00:25:9C:6B:66:EA
    Rate from WRT160N (06) to WRT160N (05): 162~216 Mbps (aprox.)
    Rate from WRT160N (05) to WRT160N (06): 150~216 Mbps (aprox.)

    - WRT160N, 00:25:9C:6B:66:EA (Floor 05), linked with WRT160N (Floor 04), 00:25:9C:BF:C6:9E
    Rate from WRT160N (05) to WRT160N (04): 300~300 Mbps (aprox.)
    Rate from WRT160N (04) to WRT160N (05): 300~300 Mbps (aprox.)

    - WRT160N, 00:25:9C:BF:C6:9E (Floor 04), linked with WRT160N (Floor 03), 00:25:9C:D0:A4:3D
    Rate from WRT160N (04) to WRT160N (03): 243~243 Mbps (aprox.)
    Rate from WRT160N (03) to WRT160N (04): 243~216 Mbps (aprox.)

    - WRT160N, 00:25:9C:D0:A4:3D (Floor 03), linked with WRT160N (Floor 02), 00:25:9C:55:E2:27
    Rate from WRT160N (03) to WRT160N (02): 162~150 Mbps (aprox.)
    Rate from WRT160N (02) to WRT160N (03): 150~108 Mbps (aprox.)

    ************************************************************************
    ** THIS IS THE PROBLEM... (note MAC Floor 01: is Cisco-Linksys)

    - WRT160N, 00:25:9C:55:E2:27 (Floor 02), linked with WRT160N (Floor 01), 68:7F:74:25:01:47
    Rate from WRT160N (02) to WRT160N (01): 11~11 Mbps (aprox.)
    Rate from WRT160N (01) to WRT160N (02): 11~2 Mbps (aprox.)

    Please, note rates... in comparison with others below
    ************************************************************************

    All MAC vendor ID are: Cisco-Linksys, acording to:
    http://standards.ieee.org/cgi-bin/ouisearch?%MAC%

    - E3000 v1
    Firmware: tomato-E3000USB-1.28.9054MIPSR2-beta-vpn3.6
    http://tomatousb.org/

    WRT160N v3 (all of them)
    Firmware: tomato-K26-1.28.9054MIPSR2-beta-Std
    http://tomatousb.org/
     
  2. gutsman7

    gutsman7 Networkin' Nut Member

    Might wanna upgrade to the latest Tomato Firmware mods check out Toastmans or Shibbys they rock!
     
  3. xdmv

    xdmv LI Guru Member

    Thanks for your answer...
    I think that could be answer, but I cannot figure out...

    - Is it a MAC incompatibility?
    - Is it because WDS has a limit chaining APs?
    - Firmware has limitations chaining APs?

    NVRAM was completly erased in each AP and reconfigured from scratch.
     
  4. Azuse

    Azuse LI Guru Member

    Well WDS by it's nature halves bandwidth between APs, which is what makes it so much more stable than repeaters. So yes, WDS has an inherent limit if successive APs have weak signals. Doesn't explain you're problem.

    I'm going to go out on a limb here and point out 11 Mbps is 802.11b speeds. Either you've got some b device connected or your wireless settings are wrong.

    If you need access for b devices then a dedicated AP wired in is the best option, however as they're limited to WEP it makes the whole network more vulnerable. On the other hand b devices are still being produced - psp being the most common - but older nintendo DS and earlier games suffer the same problems.
     
  5. xdmv

    xdmv LI Guru Member

    >gutsman7, I forgot to say thanks!

    >Azuse, thanks to you too.

    I re-checked wireless settings either in AP and 2 wireless cameras connected to it.
    They work in G mode. There is no other connected device.

    I thinks it is a problem related to MAC series...
    All WRT160N v3 are 00:25:9C MAC vendor, except last one which is WRT160N v3 with a 68:7F:74 MAC vendor.

    As a said before, I flashed both cameras and AP in this floor, but I get same results.
    Very strange...

    Could this be possible?, I mean, different MACs leads to this problem...
    Please, if anybody could confirm that.

    BTW, what is the limit when chaining AP via WDS?

    Thanks!
     
  6. xtacydima

    xtacydima LI Guru Member

    Have you tried switching Floor 1 with a higher Floor, like 4 or 5 to see if the problem suddenly exists higher in the chain?
     
  7. xdmv

    xdmv LI Guru Member

    >xtacydima, Thanks! I will try that...
     
  8. xdmv

    xdmv LI Guru Member

    Problem confirmed... The cause is the Router/AP (with different MAC series).
    I changed the floor, and I got the same result.
    Then I added another WRT160N with the same MAC series, and it worked flawless.

    Does anybody can confirm if this is a bug between different MAC series?
    Thanks!
     
  9. koitsu

    koitsu Network Guru Member

    Can't help with the problem, but per OUI listing:

    00-25-2E (hex) Cisco SPVTG​
    00252E (base 16) Cisco SPVTG​
    5030 Sugarloaf Pkwy​
    Lawrenceville GA 30044​
    UNITED STATES​

    68-7F-74 (hex) Cisco-Linksys, LLC​
    687F74 (base 16) Cisco-Linksys, LLC​
    121 Theory Drive​
    Irvine California 92612​
    UNITED STATES​

    Both Cisco, but different companies. Probably different PHYs or different revision wifi chips inside. Hardware vendors change internals all the time w/out changing the product number.

    Please provide output from the wl revinfo command (only available via CLI) on both routers which are experiencing the problem. Output from wl cap would also be useful. Please be sure to state which MAC/router has which output. Then provide the same output from a router which does not experience the problem.

    Also, I do not know why you made a brand new thread for this problem. Please don't scatter this information across multiple threads, it makes it very hard for people to follow + people searching online to find coherent information:

    http://www.linksysinfo.org/index.ph...-wrt160n-drops-performance-in-wds-mode.39097/
     
  10. xdmv

    xdmv LI Guru Member

    These are the outputs:

    WRT160N v3 / MAC: 00:25:9C
    ** GOOD AP **

    - wl revinfo
    vendorid 0x14e4
    deviceid 0x4329
    radiorev 0x92056000
    chipnum 0xb83a
    chiprev 0x0
    corerev 0x11
    boardid 0x4cd
    boardvendor 0x14e4
    boardrev P700
    driverrev 0x50a9300
    ucoderev 0x1fc009d
    bus 0x0
    phytype 0x4
    phyrev 0x6
    anarev 0x8

    - wl cap
    ap sta wet mac_spoof led wme pio 802.11d 802.11h rm cqa mbss16 afterburner ampdu amsdurx amsdutx wmf rxchain_pwrsave radio_pwrsave


    WRT160N v3 / MAC: 68-7f-74
    ** BAD AP **

    - wl revinfo
    vendorid 0x14e4
    deviceid 0x4329
    radiorev 0x92056000
    chipnum 0xb83a
    chiprev 0x0
    corerev 0x11
    boardid 0x4cd
    boardvendor 0x14e4
    boardrev P700
    driverrev 0x50a9300
    ucoderev 0x1fc009d
    bus 0x0
    phytype 0x4
    phyrev 0x6
    anarev 0x8

    - wl cap
    ap sta wet mac_spoof led wme pio 802.11d 802.11h rm cqa mbss16 afterburner ampdu amsdurx amsdutx wmf rxchain_pwrsave radio_pwrsave

    Thanks!
     
  11. koitsu

    koitsu Network Guru Member

    Underlying wifi chips are 100% identical; no difference in any way/shape/form unless it is a downright silicon-level (transistor-level) change which isn't documented (which would be a little odd; they would normally increment one of the revision IDs shown in the wl revinfo output).

    The aforementioned data (from wl cap and wl revinfo) comes directly from the underlying wifi chip (e.g. it's not a compile-time setting in the wl driver). So I have no explanation for your problem. I can assure you that the MAC address itself (meaning those first 3 octets) has nothing to do with the problem -- many companies now have multiple OUIs (same goes for companies with multiple PCI Vendor IDs), cuz in some cases they actually run out (make too many products).

    It may be possible that you simply got a bad batch of routers (possibly antennas are faulty, internal connectors/soldering may be bad, etc.). I really don't know. But from the above the two wifi chips between the "good" vs. "bad" routers are 100% identical.

    If you can confirm with absolute 100% certainty that your NVRAM variables are all identical between the two models (aside from the obvious stuff that should be different) and that you don't have any "leftover" crap in NVRAM on the "bad" routers, then I'm at a loss to explain the problem. Otherwise if you have leftover crap in NVRAM (say from trying other firmwares, or from upgrading from stock firmware to Tomato/TomatoUSB and forgetting to choose the "Erase NVRAM (thorough)" option (the thorough one is needed!)), I can imagine this could cause oddities depending on the underlying code and what sort of "legacy" NVRAM variables it may actually still honour/use.
     
  12. xdmv

    xdmv LI Guru Member

    >koitsu, Thanks a lot.
    I'm pretty sure the NVRAM clean up process was done.
    Also, they have the same firmware.
    I checked all three times.

    So here it comes the question: Is this a limit for WDS chain? I have:
    Internet <> E3000(Gateway) <> WRT160N <>WRT160N <>WRT160N <>WRT160N <>WRT160N <>WRT160N(* BAD)
    They all are chained via wireless using AP + WDS mode.
    Could this be a chain limit?
    They all work in channel 9, with no interference.

    Anyway, the whole process works good.
    I'm just only curious why this is happen.

    Thanks!
     

Share This Page