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

Tomato RAF Releases

Discussion in 'Tomato Firmware' started by Victek, Dec 28, 2012.

  1. Guzel

    Guzel Network Guru Member

    I have 1000/1000 , I get 283/181 whit E4200 Speedboost, i have trayd whit my tomato 3500l beta16 whit Fastnat i get the same 268/177
     
  2. hacim

    hacim Networkin' Nut Member

    I'm looking for hardware advice for running RAF. I want to build a AP that will *primarily* be connected via OpenVPN for all traffic. I used the E3000 in the past, but I was not so happy with it, it worked much better than my linksys WRT54G, but signal strength was not as good as I thought. Anyways, any advice would be recommended!
     
    Last edited: Oct 12, 2013
  3. Daky

    Daky Network Guru Member

    I have a couple of servers to co-locate at your place :p
     
  4. Elfew

    Elfew Addicted to LI Member

    I just ignore Nurofen - his questions are annoying.

    @Victek - thx for your answer.

    IPset is not problem no? Just modules without GUI, am I right?
     
  5. nurofen

    nurofen Serious Server Member

    @Edrikk, @Victek - Excuse me please. I did not mean that....
    I am learning with your help. I really liked your firmware :)

    @Elfew,Newbie questions, do not swear)

    Thank you.









     
  6. kthaddock

    kthaddock Network Guru Member

    He has problem with language be patient, maby he use translator.
    You have all been noob so don't be cocky.
     
    Last edited: Oct 13, 2013
    Marcel Tunks and nurofen like this.
  7. Victek

    Victek Network Guru Member

    That's good test ... taking in mind that E4200 or WNR3500l CPU is the bottleneck... What router do you use now for 1Gbps access?

    Try E3000 again, you have firmware updated to test now.

    Uh? You was talking about Exfat... ipset ... I'm not working on it now.

    @kthaddock ... don't agree with you, review all his posts using the forum tool, then edit your post if you think that the opinion of other people makes sense ;). And I answer all @nurofen questions the same as I do to any other forum visitor.
     
    Last edited: Oct 12, 2013
  8. RonV

    RonV Network Guru Member

    Victek,

    Updated to the Tomato RAF Firmware v1.28.9013 MIPSR2-RAF-V1.2m K26 USB today on my N66U. So far IPv6 appears stable but same issue as before when QOS rules are in place. IPv6 connectivity goes away and the logs show a failing ping as before. I am uploading my debugging dumps as before in a zip file in case you want to look at them. Thanks for cranking the update for the weekend, keeps me busy testing...

    Code:
    Oct 12 16:06:51 rtr-gateway-n66r user.debug kernel: icmpv6_send: no reply to icmp error
    Oct 12 16:06:51 rtr-gateway-n66r user.debug kernel: icmpv6_send: no reply to icmp error
    Oct 12 16:06:51 rtr-gateway-n66r user.debug kernel: icmpv6_send: no reply to icmp error
    
     

    Attached Files:

  9. Victek

    Victek Network Guru Member

    Hi RonV, Thank you for testing and the log, it will help me a lot cause now I have also Lorenceo logs from his dhcpv6pd connection, finally works and he was testing also QoS. I can compare.
     
  10. mercurio_

    mercurio_ Reformed Router Member

    hy victek,

    i just tried your beta build for the ac66u. it seems that the client mode only works for open and WEP encrypted networks. any idea why?
     
  11. vlads

    vlads Serious Server Member

    @Victek - thanks for the builds. This is to confirm that the WNR3500Lv2 WAN port configuration page issues have been fixed. I suspect the fix extends to the E1200v1 as well (reported by @Daky).
     
  12. Victek

    Victek Network Guru Member

    Since I have no router to test I have to believe what you said ... no I have no idea why, did you try other builds?

    WEP for a/c/n wireless networks? ... are you sure? is not supported by the 802.11ac specs... are you may be talking for wireless b/g?

    Good to know, compilation switches finally fixed for these builds. Also you see that more features are now available for WNR3500L v2 build, not?
     
    Last edited: Oct 13, 2013
  13. mercurio_

    mercurio_ Reformed Router Member

    thanks for the fast reply.

    i'm quite sure, because if i configure my e1200 v2 (running shibbys tomato 1.28) the same way everything works great. furthermore on the ac66u the bridge mode also works for WPA and WPA2.

    but the networks are not set to n or c. they are all on b/g.

    if i switch from bridge to client it stops working (when the encryption is not wep or open)

    the behavior is the same for shibbys builds.

    by the way, could you may build a e1200 v2 beta?

    thanks a lot

    if i can help you by getting a log tell me. i tried to find out what's going on having a look at the log files, but i don't really know what to look for.
     
  14. Victek

    Victek Network Guru Member

    Ok, I can't test as I said before, pity you use an AC router for b/g wireless ... about e1200 firmware, please look at the beta section.. it's.
     
  15. mercurio_

    mercurio_ Reformed Router Member

    thanks once more. i had a look in the beta section but only found e1200v1 and not e1200v2. they are somewhat different from what i know, am i wrong?

    i only use the 2.4ghz channel in b/g. is there some way i could find out why it does not connect when using wpa? i mean do you have anny advice what i could try to locate the error?
     
  16. Victek

    Victek Network Guru Member

    Do you see it now? .. Thanks!

    Screenshot from 2013-10-13 02:12:32.png

    About wpa2 psk-aes -- did you use the correct lenght for the password? The encryption key may be from 8 to 63 printable ASCII characters or 64hexadecimal digits, right?
     
  17. mercurio_

    mercurio_ Reformed Router Member

    thanks, i was looking here: http://victek.is-a-geek.com/Repositorios/tomatoraf_v1.3/

    and yes the password length is correct.
    what i noticed now is that if i change from bridge to client mode the ehternet port status gets confused somehow. it adds a port and shifts the other to the right.

    e.g.: if in bridged mode port 1 and port 4 are connected to some ethernet device everything is fine, they show the connection in the gui. once client mode is selected the guy shows only a connection on port 2 (although port 1 and 4 are connected)
    if in bridged mode port 1 and port 3 are connected to some ethernet device everything is fine, they show the connection in the gui. once client mode is selected the guy shows a connection on port 2 and one on port 4 (although port 1 and 3 are connected)

    could it be that the wifi should be mapped to the wan port (that's how tomato behaves on the e1200) and because of some error it's mapped to a new port instead?

    i also noticed that in advanced/vlan the ports are wrong mapped (port4 vlan = port1 ethernet, port3 vlan =port2 ethernet, port2 vlan=port3 ethernet, port1 vlan=port4 ethernet)
     
    Last edited: Oct 13, 2013
  18. Victek

    Victek Network Guru Member

    Better use the GUI .. is easy.

    The port mapping it's designed for device configuration as router or gateway and to verify wired speed, other modes could show erratic behavior, you better disable if you use as wireless bridge..not?.
     
  19. mercurio_

    mercurio_ Reformed Router Member

    thanks, didn't know that. one last question, is there some way to see what the wifi driver is doing. i connected trough ssh and tried to get some info using "wl -i ...". who handles the initialization of the wifi driver is there some way to access infos and warnings related to bringing up eth1?
     
  20. vlads

    vlads Serious Server Member

    I noticed the USB menu, however the web server and JFFS (really nice to have since the WNR3500Lv2 has 128MB of flash) are still not there.
     
  21. koitsu

    koitsu Network Guru Member

    Nope -- and you can blame Broadcom for that. The wireless drivers are binary blobs, so "infos and warnings" are at their discretion to printk() into the kernel log (see dmesg) or not. The Broadcom stuff is known to crap out without emitting a single line/error/issue. There is nothing anyone can do about this due to the closed-source nature of the driver. Please direct all complaints to Broadcom.

    Also, I strongly recommend you do not "fool around" with wl to see what different things do/start playing with things. Many of the features/subcommands do not work on certain chips, behave wrongly/badly on others, or conflict with other parts of the subsystem and cause issues. No, which commands on what chips/hardware behaving badly is not documented, it varies too much. Just leave the command alone/don't screw around.
     
  22. ksergey2012

    ksergey2012 Serious Server Member

    Good day to all! Dear Victek, I fell back to firmware 1.2h - everything fell into place - and WAN connected devices instantly and displayed correctly screenshot: http://yadi.sk/d/y27KYVtbAq7xV and all the feelings firmware 1.2h very stable and nimble. Thank you for the great work! By the way, about the power of the signal, a double reboot - does not solve anything, if you install Country United States, it is better to use the Country EU.
     
  23. Ric

    Ric Network Guru Member

    Appears to be routing or DNS issues with IPV6 when using 6rd. There are times that I load it up, get the IPV6 address on the router and it will propagate to the connected devices. It will work for a bit, and even then gets sluggish after some time. I can run an IPV6-Test and all checks good initially, and then will fail. I'm not certain if it's DNS or possibly a routing issue. The following is my configuration:

    1. 6rd Service Provider IPv6 prefix: 2602:300::
    2. Prefix length: 28
    3. IPv4 mask length: 0 (to allow all 32 bits to be used in the delegated prefix)
    4. IPv4 address of the Border relay: 12.83.49.81

    Currently using Release 1.2m with an E4200. This problem is not new to this release. Has been there forever. Just figured I'd mention it since there has been so much effort put into IPV6 recently.
     
  24. mercurio_

    mercurio_ Reformed Router Member

    thanks for the informative answer,

    do you may know how bringing the wifi interfaces up is done in tomato? i mean which script / program takes care of that?
    the point is that apparently the wifi comes up but either does not request a dhcp lease or does not complete the association process. now i wonder what i could do to see what's going on and may find the error. i hoped that somewhere are some logs that at least show the failed / succeeded tasks when bringing the interface up.
     
  25. Victek

    Victek Network Guru Member

    Hi, welcome to the ipv6 migration club! ;)

    Well, another candidate to improve dnsmasq bugs, let's go, interesting .. prefix /28 .. wow! a big one.. dam, Could you provide your router syslog? ... the key points that can help us to solve your config are:

    show me .... cat /proc/sys/net/ipv6/conf/br0/mtu

    and .. I know 6rd is not working in your case... simply is not deployed in tomato code, but I hope you can help us with your information, so, you're right .. doesn't work correctly.

    Code:
    switch(E('_ipv6_service').value) {
            case 'other':
                fom.ipv6_prefix_length.value = 64;
                fom.ipv6_prefix.value = '';
                fom.ipv6_rtr_addr.value = fom.f_ipv6_rtr_addr.value;
                break;
            case '6rd':
                break; //KDB todo
            case '6to4':
            case 'native-pd':
                fom.ipv6_prefix.value = '';
                fom.ipv6_rtr_addr.value = '';
                break;
            default:
                fom.ipv6_rtr_addr.disabled = fom.f_ipv6_rtr_addr_auto.disabled;
                if (fom.f_ipv6_rtr_addr_auto.value == 1)
                    fom.ipv6_rtr_addr.value = fom.f_ipv6_rtr_addr.value;
                else
                    fom.ipv6_rtr_addr.value = '';
                break;
        }
    Edit: 6rd is not implemented in kernel 2.6.22 ... so .. kernel patch again.

    Thanks!
     
    Last edited: Oct 13, 2013
  26. eviltone

    eviltone Network Guru Member

    Victek -- i cant access your site to grab 1.2m -- the site is returning "connection reset"

    EDIT -- Either you fixed it -- or me dumping my DNS cache fixed it... Sorry for being a pain....
     
  27. Victek

    Victek Network Guru Member

    I think dns caching... I'm in the cave ;)
     
    koitsu likes this.
  28. Ric

    Ric Network Guru Member

    Thanks for the response Victek! While IPV6 is not really necessary here right now, I have been trying to play with it a little just to learn more about it.

    1500

    Here's the current syslog after a short period of rebooting with 6brd enabled/configured. Was able to successfully run IPV6-test from a WIN7 system and a Galaxy S3.
     

    Attached Files:

  29. RonV

    RonV Network Guru Member

    Well with the "m" release I am getting a 10 out of 10 on readiness for IPv6:

    ipv6.png
     
  30. Victek

    Victek Network Guru Member

    Yes, I see no problems in your log surfing with ipv6, then, what's wrong?, you cannot navigate after a short period.. what period? what message did you get?... what's your dnsmasq.conf file?.

    Thanks!
     
  31. Ric

    Ric Network Guru Member

    After approx 15-min IPV6 becomes unresponsive, and IPV4 sites load slow or not at all.

    dnsmasq.conf:
     
    Last edited: Oct 13, 2013
  32. Ric

    Ric Network Guru Member

    Currently, my desktop ws received IPV6 add but Galaxy S3 and another win7 system did not.
     
  33. Victek

    Victek Network Guru Member

    I don't know if you did before... did you erased nvram and configure manually again?.
    dnsmasq.conf looks clean and normal ..please delete the contents you pasted.. just a MAC address discovery..

    I completed code for 6rd ... go to build another test version soon ... not today, I would like to test in my site... 'unfortunately' everything works in my ipv6 tunnel but nevertheless I prefer to test as usual before release.

    Thanks!
     
  34. Ric

    Ric Network Guru Member

    Every time I update firmware I do complete NVRAM erase and clean/manual configure. From previous post, took about 10 minutes for IPV6 to propagate to workstations, but it did start working.
     
  35. Victek

    Victek Network Guru Member

    What's your router model?.. just to send you the link to test changes I did..
     
  36. Ric

    Ric Network Guru Member

    E4200 v1
     
  37. Victek

    Victek Network Guru Member

    Sorry for delay, test version for E4200V1 available in Beta section, I'm very interested to receive feedback. I'll clone to other models if feedback is OK.
     
  38. mercurio_

    mercurio_ Reformed Router Member

    i do not want to spam this thread with something none is interested into, i'm sorry if i do.

    i looked somewhat more into the wires client mode. from all the posts i found around the web (also in other forums) i got the following image: client mode seems broken for a lot of dual band routers if encryption is other than wep or open.

    as i have three routers lying around (one e900, one e1200v2 and a ac66u) i did the following test:

    i used the e1200v2 as main router in ap mode.
    i used the e900 and the ac66u in wireless client / wireless bridge mode.

    all routers where fresh flashed with the last victek firmwares ( tested both, 1.3 beta and 1.2 )

    the results can be summarized in the following two situations:

    e1200 wep encrypted / open AP:

    only changed wifi interface setting to client mode and set AP name and wep key for
    *e900: everything works fine. on the status page the wan port icon blinks as soon as the router connects to the AP
    *ac66u: everything works fine. the status page port connections are shifted by one to the right and the wan port icon does not blink
    *e1200: shows the routers in the devices page with an assigned ip address

    e1200 wpa/wpa2 AP:

    as above only changed the wifi mode after flashing and clearing nvram
    *e900: everything works fine. same situation as for wep / open
    *ac66u: does not connect to the AP. ports in the status page shifted as for client in wep mode, the AP network is shown with signal strength on the status page.
    *e1200: shows the e900 router in the devices page with an assigned ip. the ac66u appears and disappears sporadically and never gets an ip.

    if i change the wifi client setting and intentionally set a worng passphrase for the wpa connection:
    *e900: does not connect to the AP, does not get an ip (yes, i know it's obvious), the AP network is shown with signal strength on the status page.
    *ac66u: behaves the same way as the e900 beside the shift of the ports in the status page.
    *e1200: behaves as mentioned above for the ac66u ( the client router appear and disappears sporadically and never get an ip)

    i don't know if this basic test helps at all to find out what's going wrong. if i could run some tests please tell me. would be great if client mode could be fixed for dual band routers. (btw. i know i could switch to dd wrt, which i used for years and where the client mode works for the ac66u, but i have one problem, i fell in love with tomato ;-) )
     
  39. Ric

    Ric Network Guru Member

    Finally, some results for you. For the 6rd link, the MTU still set at 1500. I figured since it's a PPPoE connection that the MTU would drop to 1472 for the 6rd tunnel.

    I'm seeing slow response on IPV6 connections and pages are timing out (due to MTU?).

    Odd thing I noticed was the IPV6 connect info passed to the WIN7 machine (see below):

    Ethernet adapter Local Area Connection:
    Connection-specific DNS Suffix . :
    Description . . . . . . . . . . . : Realtek PCIe GBE Family Controller
    Physical Address. . . . . . . . . : 38-60-77-D8-98-A4
    DHCP Enabled. . . . . . . . . . . : No
    Autoconfiguration Enabled . . . . : Yes
    IPv6 Address. . . . . . . . . . . : 2602:304:ab2d:9320::5c1d:3a52(Preferred)
    Lease Obtained. . . . . . . . . . : Sunday, October 13, 2013 10:55:44 PM
    Lease Expires . . . . . . . . . . : Monday, October 14, 2013 10:55:48 AM
    IPv6 Address. . . . . . . . . . . : 2602:304:ab2d:9320:df9:e6af:95c8:fbce(Preferred)
    Temporary IPv6 Address. . . . . . : 2602:304:ab2d:9320:fc70:28ba:215c:e974(Preferred)
    Link-local IPv6 Address . . . . . : fe80::df9:e6af:95c8:fbce%13(Preferred)
    IPv4 Address. . . . . . . . . . . : 192.168.1.41(Preferred)
    Subnet Mask . . . . . . . . . . . : 255.255.255.0
    Default Gateway . . . . . . . . . : fe80::5a6d:8fff:fe3e:b17%13
    192.168.1.1
    DHCPv6 IAID . . . . . . . . . . . : 272130167
    DHCPv6 Client DUID. . . . . . . . : 00-01-00-01-17-B9-E3-F7-38-60-77-D8-98-A4
    DNS Servers . . . . . . . . . . . : 2602:304:ab2d:9320::1
    192.168.1.1
    NetBIOS over Tcpip. . . . . . . . : Enabled

    Not certain if this will cause routing problems or not.

    The other thing that I noticed is that now when you look at the device screen in the router it will list some, but not all of the IPV6 addresses that are assigned to the clients. I was seeing them for 2 clients on eth1, 1 client on br0, but none for eth2.

    Wish I had more time right now to dig, but have to go to work. Appreciate all you do! -Ric
     
  40. Victek

    Victek Network Guru Member

    Thanks @Ric , answers:

    a) Are you able to ping, traceroute (tracert) or surf ipv6 sites?
    b) Some ipv6 sites responds, others not?

    In fact mtu according with the opinion of dnsmasq programmer is not the issue since the tunnel will harmonize the packet overhead... nevertheless you can change your mtu size with the following command in telnet mode:

    echo 1472> /proc/sys/net/ipv6/conf/br0/mtu

    And run the a) && b) test again.

    Attached you can see my devices-list, after the changes.. some information appear but interface is not showed in my case... working on it, lease periods are not the same now for ipv4 and ipv6 .. it's something I have to add in ipv6 settings for the constructor, now it's fixed to 12 hours.

    Screenshot from 2013-10-14 09:06:07.png


    Cheers.

    @mercurio_ thanks for your findings, unfortunately and mentioned above I can't do so many things when we 'insert' binary drivers with few manufacturer information.
     
    Last edited: Oct 14, 2013
    Elfew likes this.
  41. Elfew

    Elfew Addicted to LI Member

    Ipv6 brings so many problems with features in tomato - qos, bw limiter, acces restriction and device list. Keep good work.

    One question - is possible to implement support for ipv6 in these features and not only handle it by mac
     
  42. Victek

    Victek Network Guru Member

    I think (correct me if my lonely neuron is wrong) easier to play with MAC address since it's in the arp and it's the device footprint. I can't figure someone willing to limit ipv6 QoS and not ipv4 QoS .. if so .. then we'll face with dual GUI's in most features, not only the related by you....
    At the moment issues with ipv6 are linked with 'tempremental' or race conditions using the service, also ISP's settings add more 'fog' to the environment, it add difficulties to discover without appropriate tools (ipv6 access as the user have) and information what's going on. That's the reason why I'm releasing so many versions and asking feedback... sometimes it's a blind and feelings against a steeplechase.
     
    Last edited: Oct 14, 2013
  43. Lorenceo

    Lorenceo Networkin' Nut Member

    I've found that if I erase the NVRAM, and then enter my v6 and PPP settings, ipv6 will work on 1.2n. However, this only seems to last until the router is power cycled. Once it restarts all v6 hosts are apparently unreachable. Clients obtain their v6 addresses, the router still sends out router advertisements, obtains its v6 prefix and address, but all v6 hosts are reported as unreachable.

    Another situation in which it seems to work is after flashing a new version. Again, it only lasts up until the router is rebooted.

    I've compared the nvram between an updated, working router, and then that same rebooted router, and all values except the IPv4 and IPv6 addresses are the same.
     
  44. Elfew

    Elfew Addicted to LI Member


    Sometimes you need IP or IP range for limiting or other services... Now working ipv6 is important, then adjust qos etc.
     
  45. Victek

    Victek Network Guru Member

    It's usual with dynamic IP's .... we need a Ghostbuster ... or the router ignores some nvram value... as I'm finding with RT-AC56U nvram but I think it's related to SMP CPU...

    Lorenceo, are you sure that a 'new' nvram value was added after flashing in the nvram list or some typo in the value name? besides ... rrules_timewarn=1
     
    Last edited: Oct 14, 2013
  46. Lorenceo

    Lorenceo Networkin' Nut Member

    I just tried flashing 1.2n on my N16. It exhibits the exact same symptoms as the N66u, except that v6 didn't start working after the flash, or after erasing the NVRAM.
    Perhaps v6 working on the N66u after flashing, and after erasing NVRAM is a coincidence.

    When it wants to it works well. Unfortunately most of the time it doesn't seem to want to work.

    @Victek The NVRAM looked exactly the same. I compared the two with notepad++. The only lines which were different were:
    ipv6_prefix
    ipv6_rtr_addr
    wan_ipaddr_buf
    wan_ipaddr
     
  47. mercurio_

    mercurio_ Reformed Router Member

    @Victek :
    one more question. i was planing to have a look at the source code and compare the versions for the ac66u and the e900. but as i don't know to much about tomato it will bi quite some work.

    did i get it right that you think the problem is related to the wifi driver and not to the tomato code itself?

    i thought that as the difference shows up in the status page there is also some difference in the source and maybe a wrong routed interface and as it behaves like the e900 with wrong passkey maybe there is some error regarding the passkey.

    but if the problem is not in changes in the tomato source then there is nothing to look for.
     
  48. Victek

    Victek Network Guru Member

    @mercurio_ code is the same with minor differences related to wan-lan-wireless interface naming, routing is the same, wireless are not the same for rt-n and rt-ac routers but it's only the binaries.

    Hands up for people testing ipv6 .. please disable additional tomato modules like QoS, Access Restriction, iptraffic ... these modules are not tested to work with ipv6 yet....
     
  49. Miguel

    Miguel Reformed Router Member

    @Victek, First of all thank you for all you hard work! I have a Asus RT-AC66U and your latest firmware as well as a E4200... there is just no way to get the VoIP to work with VLAN3... routing does not show VLAN3 at all... I have been trying several times and just no luck! I would honestly like to get rid of the Mikrotik and put one of those routers. I did follow your FTTH instructions but they do not work.

    Thanks again for all you do...
     
  50. RonV

    RonV Network Guru Member


    Yes I agree...that QOS is definatly broken, Acess Restrictions seem to work unless you are using "String". IP Traffic monitor is running and looks like it's catching my traffic but I don't know if the numbers are really accurate.
     
  51. mercurio_

    mercurio_ Reformed Router Member

    thanks again for the reply. today i tested the client mode with dd-wrt and it works without any problems. do you think this is mainly because dd-wrt runs on a 3.x kernel and uses one other wl driver? or would it may possible to compare the sources to find the difference?

    by the way, i don't know if everyone nows but the wps button clears the nvram when pressed while booting up (i only knew that the reset button puts the unit in recovery mode)
     
  52. Victek

    Victek Network Guru Member

    Can't be compared... kernel, drivers ... code, but at least you found the solution !!.
    Yes, we know it ... but if you look in administration/buttons-led you have other possibilities ... you can customize.
     
    Last edited: Oct 15, 2013
  53. Victek

    Victek Network Guru Member

    Don't give up... can't compare RT-AC66U against rb-750gl, well, I changed some settings for ripd on latest builds and additional configuration is needed, have to modify screenshots in my site too, I'll do earlier, meanwhile you need to wake up vlan3 typing these commands from telnet session:

    Code:
    root@RT-AC56U:/tmp/home/root# ifconfig -a vlan3 up
    root@RT-AC56U:/tmp/home/root# udhcpc -i vlan3
    Then vlan3 should appear in advanced/routing gui... these commands are additional as the vlan3 setting using advanced/vlan gui also. These settings will be modified soon since zebra will be changed by quagga in next version....http://www.nongnu.org/quagga/ ... I have to do something to face ipv6 for these services and zebra is deprecated...
     
    Last edited: Oct 15, 2013
  54. jimmc

    jimmc Reformed Router Member

    Please excuse a simple question from a total newbie...
    The CRC-32 for the Belkin F7D4301 build appears to to be incorrect, has the file been updated to address the 32K nvram issue (posts #1467/8).
     
  55. Victek

    Victek Network Guru Member

    If you point to firmware version in Beta area (bottom view) then yes, it's a new release (1.2e)... I got feedback that nvram space was ok.
     
  56. Miguel

    Miguel Reformed Router Member

    @Victek Thanks... will try tonight (when I get home) and confirm. I think that would be very usefull for many users in Spain that want to change their standard Movistar Routers for something with better permormance and management.
     
  57. Victek

    Victek Network Guru Member

    That's the goal ...even I would like to create a Movistar Configuration button in configuration settings to go straight without making weird maneuvers .. but I'm dancing with many subjects right now.
     
  58. Ric

    Ric Network Guru Member

    a) Yes. I am able to ping and surf ipv6 sites.
    and
    b) Correct. Some sites respond, some do not.


    I ran ifconfig on the router (telnet) and the mtu for the six0 interface is set to 1472 as it should be. br0 is set to 1500

    That's exactly what I'm seeing here. Not all macs are showing there, but a few are.
     
  59. Victek

    Victek Network Guru Member

    OK, the dnsmasq programmer is doing changes every day in dnsmasq to sort these issues, http://thekelleys.org.uk/gitweb/?p=dnsmasq.git;a=summary .. we in tomato adopted dnsmasq as main package too fast while other firmware colleagues are looking for a more mature development, in any case we are here and we'll solve with your help and other users information.

    The good news is that I can build a new version late tonight including these new changes form last 5 hours, the bad news is that I'm able to surf ipv6 global without any problem due to ipv6 tunneling connection I have.

    Almost of the issues pointed in this thread related to ipv6 come from users having 'exotic' connections like your case .... do you have 2^128 − 28 = 2^100 .. addresses .... simply crazy... are you the main ISP in your country? ;) .. I know, you're one user having ipv6 from a 'funny' ISP ...
     
    Last edited: Oct 15, 2013
  60. ksergey2012

    ksergey2012 Serious Server Member

    Dear Victek, decided yet on the test firmware 1.2m for asus RT-AC66U. All the same, it is in the firmware - not so (In firmware error). IP - address (which writes the router) for WAN connections do not correspond to the reality, the screenshot: http://yadi.sk/d/5zaj-apkAzrQy http://yadi.sk/d/7dgfQNHwAzrhq And the NAS, which is connected to the router also determines the external IP - address that identifies Speedtest, screenshot: http://yadi.sk/d/-fVsp8FeAzt3z This, of course, not possible for the host name (ddns) to remotely connect to the NAS. Perhaps that is why in the new firmware has a long and WAN connected? In firmware 1.2h - was no such problem.
     
  61. Victek

    Victek Network Guru Member

    ksergey, thanks for the information, I explain what's going on in your case, it's a recent issue I discovered:
    The IP shown in the GUI is the IP stored in /etc/hosts file (in your router). This IP is the first one written when your router using pppoe establish a link between the router and the ISP concentrator, it's clear until this point? Yes? I follow with the interesting part now:

    Then for unknown reasons your ISP broke the link, then your router dial up again and obtained a new public IP .. but .. this IP is not passed to /etc/hosts file because the communication between the pppoe daemon and the pppd module doesn't exists (it's a complete zombie module) in tomato latest versions. This is the reason why the GUI shows this 'old' IP.

    Yes.. WE have to investigate how to solve it but I can say you that no changes have been done from 1.2h version till now, I figure that your ISP changed something or they are doing funny things. Would you try again version 1.2h and confirm that 1.2h is working fine? It will help me a lot also and will stop everything I'm doing now to check this subject ... free of charge .. as always.

    Thanks! ;)
     
  62. ksergey2012

    ksergey2012 Serious Server Member

    Well, I'll be back on the firmware 1.2h and accomplish your goal.
     
  63. Miguel

    Miguel Reformed Router Member

    No luck... tried the telnet and seem to work but nothing shows in Advance/Routing. I tried again in both RT-AC66U and E4200...

    This is the terminal session from the E4200... result is the same in the AC66U

    Tomato v1.28.9013 MIPSR2-RAF-V1.2m K26 USB
    root@unknown:/tmp/home/root# ifconfig -a vlan3 up
    root@unknown:/tmp/home/root# udhcpc -i vlan3
    udhcpc (v1.21.1) started
    Sending discover...
    Sending select for 10.28.42.210...
    Lease of 10.28.42.210 obtained, lease time 1800
    root@unknown:/tmp/home/root#

    This was on the logs:
    Oct 15 20:09:20 unknown user.debug kernel: vlan3: add 33:33:00:00:00:02 mcast address to master interface
    Oct 15 20:09:20 unknown user.debug kernel: vlan3: add 33:33:00:00:00:01 mcast address to master interface
    Oct 15 20:09:20 unknown user.debug kernel: vlan3: add 01:00:5e:00:00:01 mcast address to master interface
     
  64. ksergey2012

    ksergey2012 Serious Server Member

    Dear Victek, on firmware 1.2h - everything works fine, no problems, screenshots: http://yadi.sk/d/uPhkZmWMB2CRT , http://yadi.sk/d/FOKgztMnB2CZh , http://yadi.sk/d/6nPe6zdUB2Cmo . I rebooted the router several times to make sure everything is okay. And, dear Victek note how quickly WAN connect to the Internet.
     
  65. Victek

    Victek Network Guru Member

    Look for settings in pppoe from 1.2h to 1.2m ... I think was changed. Did you erase nvram and reconfigure again when you downgraded to 1.2h? the timeshift betwwen update and total update is around 40 seconds?.. well .. depending on ISP you can see big variations but what's important now is to check the stock settings for 1.2h and clone to 1.2m.

    Previously I made a mistake, ipv4 public IP is keept in nvram, /etc/host keep ipv6 address...

    @Miguel, I'll check ... but the settings for vlan3 can be found typing ifconfig from telnet to check that interface is up and got public IP...
     
  66. Ric

    Ric Network Guru Member


    AT&T here. so one of the larger ISP's I believe.
     
  67. ksergey2012

    ksergey2012 Serious Server Member

    Dear Victek, I just rewrite - country, time, prescribed in the basic settings of provider data (login and password) and the encryption key - and all the other settings I use - by default. I always cleanse NVRAM - before and after the firmware. And tune settings - by hand (especially since very few of them). I immediately after the release of firmware 1.2m, said that vney that something is wrong.
     
  68. zekev

    zekev Reformed Router Member

    Hi guys, its my first post here. I have a netgear N600 aka wndr3700v3. My main goal of hacking this router is to have a better USB speed for my hard drive plugged in my router. I'v played with the dd-wrt build (I now have dd-wrt.v24-22118_NEWD-2_K3.x_mega-nv64k.bin) but no luck, my huge 3Tb HDD is not properly recognized.

    I found this forum yeah... but I'm wondering if someone can tell me if a 3Tb HDD will work with this Tomato RAF 1.28.9013.

    If yes, does someone monitored the USB speed read/write ?

    Thanks !
     
  69. kyrios

    kyrios Serious Server Member

    @ksergey2012
    Is it possible you match Redial Period of 1.2h and 1.2m to 40s or more?

    I found I was always fine setting 30s for 1.2g and below.
    For 1.2h, setting 30s will never work. PPPoE connection is never established.
    I then lengthen to 35s, and it's fine.

    I'm afraid 1.2m need at least 40s for Redial Period
     
  70. ksergey2012

    ksergey2012 Serious Server Member

    Good day to all! Dear Kyrios, I tried to set different values Redial Period, but it does nothing - the result is zero. But on firmware 1.2h, WAN when the load is connected router (the same as on the original firmware from asus). Dear Victek, made a new firmware 1.2n, I come home I will test.
     
  71. Victek

    Victek Network Guru Member

    Welcome zekev, let's wait if someone tested USB speed for this version.
     
  72. lefty

    lefty Networkin' Nut Member

    Happen to be Uverse?
     
  73. RonV

    RonV Network Guru Member

    Rick, Are you AT&T U-Verse or DSL? What model of equipment did AT&T provide? I use U-Verse with their 45 mbps service. Working well with Victeck's.
     
  74. ksergey2012

    ksergey2012 Serious Server Member

    Dear Victek, 1.2n tested the firmware for the router asus RT-AC66U. External IP - address of the router determines the correct, but the problem with a long (more than 3 minutes) connecting the WAN, and has remained (as on firmware 1.2 m). In general, the firmware is stable and fast. Thank you very much for your great work!
     
  75. kyrios

    kyrios Serious Server Member

    1.2n3 ksergey?
     
  76. McDougz

    McDougz Reformed Router Member

    Hi everyone,
    I'm a bit of a newb. I am using a E3000 router and I want to make sure my settings are optimized. I am getting a low download and upload when connected through my router. My firmware build is 1.28.9013 -- can you possibly point me to where I can see the optimal settings for this router?

    I've hardwired my connection to my laptop and I receive a download speed of approximately 24Mbps. But when I'm on wifi over 2.4Ghz with the E3000 I'm sitting around 9Mbps. The 5.0Ghz is sitting around 1Mbps. I've tried changing the beacon interval from 100 to 75. I've tried a few other small tweaks. But nothing seems to work.

    I was hopeful that upgrading to TomatoUSB might help the speed issue. Same speeds as before. Is the router pooched? I'm not opposed to buying a new router if someone can recommend a good one. But shouldn't the E3000 do the job? Laptop is in same room as router - 6 feet away.

    Thanks,
    Allan
     
    Last edited: Oct 16, 2013
  77. ksergey2012

    ksergey2012 Serious Server Member

  78. Victek

    Victek Network Guru Member

    :confused: it drives me crazy @ksergey2012 ;) you know what I did? .. simply renamed the version with letter 'n', sorry to be unfair but I was sure that the code was not the problem ... I faced similar problems with long dial-up periods due to line quality (in my case it takes less than 1 second to connect as I'll show you on lines between asterisk).....
    The syslog 'redial' is cause I did service wan restart from command to test.

    I'm happy it works but as you see every connection is a different world...

    Code:
    Jan  1 01:00:12 RT-N16 daemon.info pppd[490]: Plugin rp-pppoe.so loaded.
    Jan  1 01:00:12 RT-N16 daemon.info pppd[490]: RP-PPPoE plugin version 3.11 compiled against pppd 2.4.5
    ** Jan  1 01:00:12 RT-N16 user.info redial[492]: Started. Time: 10
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: pppd 2.4.5 started by root, uid 0
    Jan  1 01:00:12 RT-N16 daemon.info pppd[491]: PPP session is 4014 (0xfae)
    Jan  1 01:00:12 RT-N16 daemon.warn pppd[491]: Connected to 00:30:88:19:ab:8a via interface vlan6
    Jan  1 01:00:12 RT-N16 daemon.info pppd[491]: Using interface ppp0
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: Connect: ppp0 <--> vlan6
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: PAP authentication succeeded
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: peer from calling number 00:30:88:19:AB:8A authorized
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: local  IP address xx.xxx.125.233
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: remote IP address xx.xx.67.147
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: primary  DNS address xx.xx.61.250
    Jan  1 01:00:12 RT-N16 daemon.notice pppd[491]: secondary DNS address xx.xx.61.254
    ** Jan  1 01:00:12 RT-N16 user.debug ip-up[510]: 182: pptp peerdns disabled
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq[514]: started, version UNKNOWN cachesize 1500
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq[514]: compile time options: IPv6 GNU-getopt no-RTC no-DBus no-i18n no-IDN DHCP DHCPv6 no-Lua TFTP no-conntrack ipset Tomato-helper auth
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq[514]: asynchronous logging enabled, queue limit is 5 messages
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq-dhcp[514]: DHCP, IP range 10.10.2.2 -- 10.10.2.52, lease time 1d
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq-dhcp[514]: DHCPv6, IP range ::1 -- ::255.255.255.255, lease time 12h, template for br0
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq-dhcp[514]: DHCPv4-derived IPv6 names on br0
    Jan  1 01:00:12 RT-N16 daemon.info dnsmasq-dhcp[514]: router advertisement on br0
     
    Last edited: Oct 16, 2013
  79. ksergey2012

    ksergey2012 Serious Server Member

    God knows, maybe the problem as a line, but I'm sure of one thing, if I now will set your firmware 1.2h or original firmware from Asus, WAN will connect - instantly. Dear Victek, please see my file Logs: http://yadi.sk/d/d4NCSARdB5nbg Somehow, after the router is booted in the journal each hour is only one entry - MARK, MARK... And not any more records does not occur, although in the house a lot of devices that are constantly connected to the router, then turn off. But the entries in the log - no.:(
     
  80. vlads

    vlads Serious Server Member

    @ksergey2012 - do you happen to use Rostelecom as your ISP?

    Your log file shows a reply specific to Rostelecom: Remote message: % Sessions limit exceeded.
     
  81. Victek

    Victek Network Guru Member

    Well, after reading all your syslog I can tell you somethings, vlads noticed before also.. but it seems that you share your connection with others ? ... ;) or you fixed the redial time too short and sessions are still opened at the ISP side... it's pity because a correct programming of your ISP will replace the '%' by the number of open sessions...

    Code:
    Jan 1 03:00:37 unknown daemon.info pppd[598]: Using interface ppp0
    Jan 1 03:00:37 unknown daemon.notice pppd[598]: Connect: ppp0 <--> vlan2
    Jan 1 03:00:37 unknown daemon.info pppd[598]: Remote message: % Sessions limit exceeded.
    Jan 1 03:00:37 unknown daemon.err pppd[598]: PAP authentication failed
    Jan 1 03:00:37 unknown daemon.notice pppd[598]: Connection terminated.

    To see all devices activity go to dns GUI and disable the quiet checkboxes then you will see the parrot ;) and if you are worried go to administration and set debuggind (console) to 8 and you will read everything happening in the router...

    upload_2013-10-17_0-46-6.png
     
    Last edited: Oct 16, 2013
  82. Victek

    Victek Network Guru Member

    I think you can improve it with a cheaper wireless USB adaptor.. no router problem, the settings are tunned already in the default configuration, please erase nvram from administration/configuration and restore defaults erasing nvram (it's a dropbox list).
    With some wireless adapters you may restrict the band to 20MHz (specially on 2.4GHz band), it improves the result. The 1Mbps rate in 5GHz band surprises.. this is the reason why I think there is some incompatibility between router and wireless card in your laptop.
     
  83. vlads

    vlads Serious Server Member

    Rostelecom seem to run a sluggish radius backend database. The internet is full of customers complaining about disconnected sessions taking hours to clear.

    A similar issue is experienced by Teksavvy DSL cusomers with static IPs, except in this case it's done intentionally to allow routing tables to refresh. Changing the PPPoE holdoff to 60 or above fixes the issue.
    Code:
    Oct 16 17:09:35 test1 daemon.info pppd[4128]: Remote message: duplicate address detected
    
     
  84. kyrios

    kyrios Serious Server Member

    Holy c*ap!
    So the ksergey's bug is actually not a bug?
    He cheat the ISP by trying double login?

    Victek, have you messed-up something with rp-ppoe after 1.2h build?
    You added rubbish code just to accomodate the "ksergey's bug" ?
    I hope you do not add rubbish code after 1.2h build (1.2i 1.2j 1.2k.... 1.2n3)
    just to satisfy "ksergey's bug".
     
  85. ksergey2012

    ksergey2012 Serious Server Member

    Yes, a provider of "Rostelecom", which provides Internet via fiber-optic cable (technology FTTB). That so bad?
     
  86. ksergey2012

    ksergey2012 Serious Server Member

    I wonder why the message - about double the address? I'm trying to say that after the firmware 1.2h, something it has changed (not in a good way for me).
     
  87. vlads

    vlads Serious Server Member

    I'm not in a position to rate their service, all I can say is that you're not the only one having this issue.

    I suggest you change the Redial Period to 60 or even 120 seconds and give it another try.
     
  88. RMerlin

    RMerlin Network Guru Member

    If you have a modem that can also be used as a router, make sure you don't have your login configured in the modem (in addition to your external router).
     
  89. ksergey2012

    ksergey2012 Serious Server Member

    Okay, I'll try.
     
  90. ksergey2012

    ksergey2012 Serious Server Member

    I do not have a modem, I have a wire is inserted directly into a router.
     
  91. ksergey2012

    ksergey2012 Serious Server Member

    As I said before, changing the "Redial Period" practically does not change anything: "Redial Period" = 60 - WAN connected via 1m35s after clicking "Continue"; "Redial Period" = 120 - WAN connected via 3m15s.:mad:
     
  92. nurofen

    nurofen Serious Server Member

    I also use Rostelecom(technology FTTB) ppoe such errors, no! Tomato RAF Firmware v1.28.9013 MIPSR2-RAF-V1.2n K26 USB asus RT-N66U :)
    "Redial Period" = 10


     
  93. ksergey2012

    ksergey2012 Serious Server Member

    The fact of the matter is that on firmware 1.2h with the parameter "Redial Period" = 10 when you press the "Continue" WAN already - is connected, I just want to figure out what the problem is.
     
  94. Victek

    Victek Network Guru Member

    As you see there are some interesting information in the latest posts, in fact and also answering @kyrios nothing has been changed in rp-pppoe since last 2 months when it was upgraded to 1.11 release, changes since then are related with dnsmasq and other models integration. So, what's different between 1.2h and actual version? dnsmasq to harmonize compatibility with DHCPv6PD settings.
    Also it's interesting ... @nurofen use the same ISP and his settings work fine with stock settings, so, I can figure the issue can be related to your local node. Definitively I'll close the ticket for dial-up problem in my version.

    Thanks, we follow :)
     
  95. Elfew

    Elfew Addicted to LI Member

    Shibby wrote in his topic that rp-pppoe v1.11 is unstable and in next version will be replaced by 1.10... I dont know but I dont have any problems on my all units or units from friends...
     
  96. Victek

    Victek Network Guru Member

    Well.. as far I checked now in the git shibby is not using rp-ppoe 1.11 .... may be he didn't update it.
     
  97. kyrios

    kyrios Serious Server Member

  98. Victek

    Victek Network Guru Member

    I see.. he's talking about 3g modem disconnects.. anyway..I'll check Penguin guys for more information. Thanks for the hands up.;)
    But I would like to remind you that pppoe is layer2 implementation..so all handling for protocols is made by pppd ...

    Edit: Just tested rp-pppoe 3.10 in mine.. I don't see any difference in my connection, it works same as before.
     
    Last edited: Oct 17, 2013
  99. vlads

    vlads Serious Server Member

    @Victek - ksergey may be on to something - the problem seems to be with the recent dnsmasq upgrade.

    I disconnected the WAN on a a E2000 with firmware 1.2m took 197 seconds to complete the boot process.
    I reconnected the WAN and it only took 94 seconds to boot.
    I rolled back the firmware to 1.2a and the boot cycle (with the WAN disconnected) took 92 seconds.
     
  100. Victek

    Victek Network Guru Member

    This is no reliable data, do the following:
    write in a sheet update time and connected time:
    do:
    root@RT-N16:/tmp/home/root# service wan restart
    .............................................................................................................................................
    Done.
    and then check difference again: in my case 12 seconds. ;) .. it's ISP dependent .. Really .. I appreciate very much the comments and the help you provide in the thread to solve one situation that affects to ONE user, please be realistic and fly over the trees ... thanks.

    Now you think the problem is in dnsmasq? what evidences do you have to say it?... I f I say that I modified kernel also then problem will be kernel? if I say I included Exfat then it will be Exfat... OK .. :)
     
    Last edited: Oct 17, 2013

Share This Page