[Click] for help on Roofnet

欧阳鑫 yuanmmsip at gmail.com
Sun Jan 16 04:33:02 EST 2011


Hi, everyone.

     I am running roofnet experiments on 2 Ubuntu hosts with 2.6.32-26
kernel. Both are AR5416 cards with ath9k driver. All cards run in Monitor
mode.

     The problem is that hosts seems to be running under 1Mbps rate
ONLY. Command "read srcr/es.bcast_stats" shows:

200 Read handler 'srcr/es.bcast_stats' OK
DATA 363
5.167.1.211 00-25-86-A7-01-D3 seq 1295164746 period 30000 tau 300000 sent 12
last_rx 1.762133
5.167.1.211 00-25-86-A7-01-D3 [ 2 60 100 100 198 0 ]
5.167.1.211 00-25-86-A7-01-D3 [ 2 1500 100 100 197 0 ]
5.167.1.211 00-25-86-A7-01-D3 [ 4 1500 0 100 198 0 ]
5.167.1.211 00-25-86-A7-01-D3 [ 11 1500 0 100 196 0 ]
5.167.1.211 00-25-86-A7-01-D3 [ 22 1500 0 100 197 0 ]

    I do not  know why *fwd_rate* with 2Mb/s, 5.5 Mb/s and 11 Mb/s
are all 0s.

    I forced the data rate to be 5.5Mbps by *SetTXRate(11)* on PC1, then
ping PC1 on PC2, then I run tcpdump on PC2, the result is as follows:

tcpdump -i wlan0 -n -e -xX

15:59:49.476661 5.5 Mb/s 63dBm tx power [0x00000010] DA:ff:ff:ff:ff:ff:ff
SA:PC2-MAC  BSSID:00:00:00:00:00:00 LLC, dsap SNAP (0xaa) Individual, ssap
SNAP (0xaa) Command, ctrl 0x03: oui Ethernet (0x000000), ethertype Unknown
(0x0941): Unnumbered, ui, Flags [Command], length 1494
        0x0000:  aaaa 0300 0000 0941 0200 87e4 3c00 0b00  .......A....<...
        0x0010:  dc05 0000 05a7 01d3 0100 0000 75a5 324d  ............u.2M
        0x0020:  3075 0000 e093 0400                      0u......
15:59:49.476661 1.0 Mb/s [0x0000000f] DA:ff:ff:ff:ff:ff:ff SA:PC2-MAC
 BSSID:00:00:00:00:00:00 LLC, dsap SNAP (0xaa) Individual, ssap SNAP (0xaa)
Command, ctrl 0x03: oui Ethernet (0x000000), ethertype Unknown (0x0941):
Unnumbered, ui, Flags [Command], length 1494
        0x0000:  aaaa 0300 0000 0941 0200 87e4 3c00 0b00  .......A....<...
        0x0010:  dc05 0000 05a7 01d3 0100 0000 75a5 324d  ............u.2M
        0x0020:  3075 0000 e093 04                         0u.....

 15:59:48.460612 5.5 Mb/s 63dBm tx power [0x00000010] DA:PC1-MAC SA:PC2-MAC
BSSID:00:00:00:00:00:00 LLC, dsap SNAP (0xaa) Individual, ssap SNAP (0xaa)
Command, ctrl 0x03: oui Ethernet (0x000000), ethertype Unknown (0x0943):
Unnumbered, ui, Flags [Command], length 162
        0x0000:  aaaa 0300 0000 0943 0b04 0101 0000 fc7b  .......C.......{
        0x0010:  0000 0054 0000 0000 0000 0000 0000 0000 ...T............
        0x0020:  0000 0000 0000
0000                                 ........
15:59:48.460612 1.0 Mb/s [0x0000000f] DA:PC1-MAC  SA:PC2-MAC
BSSID:00:00:00:00:00:00 LLC, dsap SNAP (0xaa) Individual, ssap SNAP (0xaa)
Command, ctrl 0x03: oui Ethernet (0x000000), ethertype Unknown (0x0943):
Unnumbered, ui, Flags [Command], length 162
        0x0000:  aaaa 0300 0000 0943 0b04 0101 0000 fc7b
.......C.......{
        0x0010:  0000 0054 0000 0000 0000 0000 0000
0000 ...T............
        0x0020:  0000 0000 0000
00                                    .......

  It seemed that each packet was caught by tcpdump 2 times, and the 2
copies share the same timestamp but different rates.

  My second question is why each packet was caught with a rate of 5.5Mb/s
first, and then with a rate of 1Mb/s.

  The attachment is my configuration file: gen_config_roofnet.pl. I do not
know if there is some error with the configuration file. Would you help me
check for it?

   Thank you very much.




>
> 在 2011年1月10日 下午10:53,Roberto Riggio <roberto.riggio at create-net.org>写道:
>
>
>> Il 10/01/2011 09:38, 欧阳鑫 ha scritto:
>>
>> > Thank you very much. I have done the experiment for long time, but
>> always
>> > meet some questions.
>> > If the official mac80211modules do not support the rate and the retries
>> > radiotap fields, the new Atheros wireless cards drived by ath9k cant't
>> work
>> > at Rate 1 and 2,  does this have some relation with the
>> > reason(mac80211modules do not support the rate and the retries radiotap
>> > fields)?
>>
>> The attached script is actually using the ExtraEncap/Decal elements,
>> while it looks
>> like that the wireless interface you are using is working in adhoc mode.
>> Roofnet
>> runs over monitor interfaces. This could be one of the problems.
>>
>> > The attachment is my configuration file: gen_config_roofnet.pl. Can you
>> help
>> > me check up if there is some error?
>>
>>
>> Try to:
>> 1) run the script over an interface operating in monitor mode (created
>> using the iw
>> command)
>> 2) modify the script to use the radiotapencap/decap elements
>>
>>   > Thank you very much!
>> >
>> > 2011/1/10 Roberto Riggio <roberto.riggio at create-net.org>
>> >
>> >> Il 09/01/2011 12:39, 欧阳鑫 ha scritto:
>> >>> Hi everyone,
>> >>>       We are running experiments with Roofnet on PCs. I have
>> downloaded
>> >>> click-1.8.0, and configured, make, make install it successfully. Then
>> >>> I run: "./gen_config_roofnet.pl --dev wlan0 | click -" on TWO PCs.
>> The
>> >>> problem is, we cannot ping each other on the 2 PCs.  We do not have
>> >>> the old Atheros wireless cards, but use the new Atheros wireless cards
>> >>> (802.11n supported) drived by ath9k. I am not sure if this is ok.
>> >> it should work. However as far as I remember the official mac80211
>> >> modules do not support the rate and the retries radiotap fields. I
>> >> assume that in this case you are always transmitting at the base rate.
>> >>> The srcr interface's HWaddr is
>> >> Yes
>> >>> On PC1, "./read_handler srcr/es.bcast_stats" shows
>> >>>
>> >>> 200 Read handler 'srcr/es.bcast_stats' OK
>> >>> DATA 360
>> >>> 5.32.234.144 00-25-86-20-EA-90 seq 1294477713 period 30000 tau 300000
>> >> sent
>> >>> 45 last_rx 10.843367
>> >>> 5.32.234.144 00-25-86-20-EA-90 [ 2 60 0 0 -1 -1 ]
>> >>> 5.32.234.144 00-25-86-20-EA-90 [ 2 1500 0 0 -1 -1 ]
>> >>> 5.32.234.144 00-25-86-20-EA-90 [ 4 1500 0 0 -1 -1 ]
>> >>> 5.32.234.144 00-25-86-20-EA-90 [ 11 1500 100 100 0 0 ]
>> >>> 5.32.234.144 00-25-86-20-EA-90 [ 22 1500 100 100 0 0 ]
>> >> this is strange because it looks like you are transmitting and
>> receiving
>> >> probes at both 5.5 and 11 Mbps but not at the lower rates.
>> >>> On PC2, "./read_handler srcr/es.bcast_stats":
>> >>>
>> >>> 200 Read handler 'srcr/es.bcast_stats' OK
>> >>> DATA 352
>> >>> 5.167.1.211 00-25-86-A7-01-D3 seq 1294477686 period 30000 tau 300000
>> sent
>> >> 36
>> >>> last_rx 3.115970
>> >>> 5.167.1.211 00-25-86-A7-01-D3 [ 2 60 0 100 0 0 ]
>> >>> 5.167.1.211 00-25-86-A7-01-D3 [ 2 1500 0 100 0 0 ]
>> >>> 5.167.1.211 00-25-86-A7-01-D3 [ 4 1500 0 100 0 0 ]
>> >>> 5.167.1.211 00-25-86-A7-01-D3 [ 11 1500 100 100 0 0 ]
>> >>> 5.167.1.211 00-25-86-A7-01-D3 [ 22 1500 85 100 0 0 ]
>> >> this means that you are receiving probes for every rate but you
>> >> are successfully transmitting only at the 5.5. and 11 rates
>> >>> The questions are:
>> >>> (1) Why as neighbors, PC1 and PC2 cannot ping each other?
>> >> because you have no routes and you have no routes because
>> >> the probes at the lowest rate (1 Mbps) are not transmitted
>> >>>   (2) srcr/lt.routes shows no data, Is this OK?
>> >> no
>> >>> (3) Each PC has 3 interfaces, srcr, rate and wlan0. Which one should
>> >>> be used to do experiments (like COPE)?
>> >> srcr
>> >>> _______________________________________________
>> >>> click mailing list
>> >>> click at amsterdam.lcs.mit.edu
>> >>> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>> >>
>> >> _______________________________________________
>> >> click mailing list
>> >> click at amsterdam.lcs.mit.edu
>> >> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> click mailing list
>> >> click at amsterdam.lcs.mit.edu
>> >> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>>
>> _______________________________________________
>> click mailing list
>> click at amsterdam.lcs.mit.edu
>> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>>
>>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gen_config_roofnet.pl
Type: application/octet-stream
Size: 10937 bytes
Desc: not available
Url : http://amsterdam.lcs.mit.edu/pipermail/click/attachments/20110116/6548acaa/attachment.obj 


More information about the click mailing list