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

Tomato Shibby's Releases

Discussion in 'Tomato Firmware' started by shibby20, Feb 26, 2011.

  1. SNR

    SNR Networkin' Nut Member

  2. Elfew

    Elfew Addicted to LI Member

    just uncheck inverted ports! ;)
     
  3. leandroong

    leandroong Addicted to LI Member

    Inverted ports match my port labels.
     
  4. occamsrazor

    occamsrazor Network Guru Member

    I still think it would be better if we could just set manual labels for each port, instead of 1,2,3,4 etc. That way we could label them with the device name "Mac desktop", "VOIP box", "Movie server" etc.
     
    SNR likes this.
  5. lwk523

    lwk523 Reformed Router Member

    Hey guy, I'm new here and I must thank to shibby first for creating a awesome firmware for us. Btw, I have a question. May I have the 3G modem support list ? Is that possible to install http://www.draisberghof.de/usb_modeswitch/ into this firmware ?

    Thank in advance
    Warmest Regards
    SL
     
  6. SNR

    SNR Networkin' Nut Member

    The screenshot I've posted here is with "Invert Ports Order" unchecked (the default setting). If I check that box, then the port connection order is reversed (i.e. it becomes a mirror image of the actual connections on the back of the router) -- but the port label order does not change to follow the ports.

    IIRC this was not a problem in build 106, but I really don't want to revert to that build just to confirm this.

    At this point I'm not really sure what the intent was with this checkbox - was it to reverse the displayed port order, or the displayed label order, or both? All I know is that with the box checked or unchecked, the labels displayed on the page do not match the order of the actual labels on the back of my RT-16N.

    I do like occamsrazor's suggestion to let the labels become free-form fields. But that is a feature request, not a request for bug fix.
     
  7. SNR

    SNR Networkin' Nut Member

    Well here's a little mystery... it seems that the actual embossed markings on the LAN ports of my RT-16N do not match the order in ASUS' documentation!

    Here's a link to the latest North American Quick Start Guide at ASUS' U.S. website:

    http://dlcdnet.asus.com/pub/ASUS/wireless/RT-N16/NA7709_RT_N16_QSG.pdf

    Page 4 clearly shows the LAN ports on the back labeled 1 to 4 from left to right. The English User Manual from the same site also has the same image on page 5.

    But I'm absolutely certain that my own RT-16N has the embossed LAN port markings in reverse order, i.e. 1 to 4 from right to left (I just checked this one more time before making this post). In fact, if you trace the port labels from the back of my router over the top and to the front, they correlate with the labels on those hideously bright blue LEDs on the front. In other words, the port labels on the back, when viewed from the back, are in reverse order compared to the LEDs on the front, when viewed from the front. This makes a bit of sense when considering the possible component layout inside the case (edit: nope, no expected correlation at all, once you take a look at photos of the inside).

    For what it's worth, my router is Hardware Version A1. I'd be curious to know if anyone has an RT-16N with the LAN port order 1 to 4 from left to right (per the manuals), and if so, what Hardware Version number do you have? Thanks folks...

    EDIT: To be clear, in all of these posts I have been referring to port label order as displayed on the back panel where the actual ports are located (and labelled). The port LEDs on the front are indeed labelled 1 to 4 from left to right (on my router at least). But (I assume) that's not the part of the router that the Ethernet Ports State is intended to represent.
     
  8. SNR

    SNR Networkin' Nut Member

    So the fundamental question is: On the RT-16N, what exactly is the Ethernet Ports State intended to represent? The back panel where the RJ-45 connections are located? or the front panel where the port LEDs are located? In spite of the page graphic, the RT-16N doesn't have any green indicators at the RJ-45 connectors...
     
  9. JugsteR

    JugsteR Serious Server Member

    Free form port assignation with a decent default would solve this problem for you and every other router no?

    Signature? Signature!
     
    SNR likes this.
  10. SNR

    SNR Networkin' Nut Member

    @JugsteR, Yes it would. But to be more generic, those little green link/activity indicators should be removed from the graphic, since not all router models have those LEDs present at the RJ-45 jacks (the RT-16N does not). Also, some router models -- like the RT-16N -- actually have the lock tabs on the bottom, not the top. :mad:

    As much as I like the Ethernet Ports State display, it sure has turned out to be a can o' worms, huh? :eek:
     
  11. SNR

    SNR Networkin' Nut Member

    That brings up one more point: "Invert Ports Order" is probably not the best wording for that checkbox, as it seems to imply a result that would have the graphic rotated 180 degrees, with the lock tab notches at the bottom instead of the top. Probably better to say "Reverse Ports Order".

    On the other hand, with all the possibilities, we might need an actual "Invert Ports Display" checkbox as well... :confused:
     
  12. Lamoboy

    Lamoboy Serious Server Member

    I have installed Shibby's tomato-K26USB-1.28.RT-MIPSR2-108-Big-VPN on to my Linksys WRT610N v2 and all is working fine except 5 GHz wireless. It was working right after initial settings, but after I have change some settings that is does not apply to wireless, it's stops working. What should I do to solve this issue?
     
  13. Elfew

    Elfew Addicted to LI Member

    There is no problem in firmware, but with the case from rt-16n router... Ports are labeled bad by ASUS
     
  14. GhaladReam

    GhaladReam Network Guru Member

    Shibby.. Did the multiSSID MAC filtering issue ever get looked at? I have to run the following command to erase the mac filters for wl0.1 every time i change them, because of the counter-intuitive "fix" that teaman put in a while back to propegate the mac filters to the virtual interfaces as well I've been unsuccessful in contacting teaman about this:

    nvram unset wl0.1_macmode
    nvram unset wl0.1_maclist
    nvram commit
    reboot
     
  15. jerrm

    jerrm Network Guru Member

    Isn't this getting a little too nitpicky? Do we really want Shibby spending any more time on such a minor cosmetic feature? As long as I can see that the port labeled "LAN 1/LAN 2/etc" is connected, I can accomodate for the fact that the graphic doesn't match the left to right order. Should he also take time with an alternate vertical layout to match my Belkin routers?

    Even if it's only a 10 minute fix, I'd rather that 10 minutes go toward the WNDR3700 or the AC66U, or xt_string, or anything else functional.
     
    Toastman, vlads, Fredrik and 4 others like this.
  16. SNR

    SNR Networkin' Nut Member

    Probably not. I was just making a point that WYSI not always WYG. In this case I have no idea what I'm getting. Is the ASUS manual incorrect, or the embossed port labels on the back of the router? And how could the Ethernet Ports State display be made to accommodate conflicting data like this? And even with no conflicts on ASUS' part, different vendors do things differently with regards to port LEDs, lock tab orientation, etc.

    I agree, fix more important things first.

    Carry on... :)
     
  17. SNR

    SNR Networkin' Nut Member

    Yes, I'm inclined to believe that's the case, since the firmware should be pulling interface numbers from the hardware. Hope the case designer got replaced with someone more deserving of the job...

    --- EDIT ---

    Sorry, that wasn't fair to the case designer... :oops:

    I've checked my RT-16N one more time. I have two cables plugged into the ports with embossed labeling (right-to-left) "LAN1" and "LAN2". The blue LEDs on the front of the router confirm that there are connections on LAN 1 and LAN 2. And syslog has these startup records for "port 1(vlan1)" and "port 2(eth1)", with no other port numbers mentioned:
    Code:
    Dec 31 16:00:08 FreeWilly user.warn kernel: Algorithmics/MIPS FPU Emulator v1.5
    Dec 31 16:00:08 FreeWilly user.debug kernel: vlan1: add 33:33:00:00:00:01 mcast address to master interface
    Dec 31 16:00:08 FreeWilly user.debug kernel: vlan1: add 01:00:5e:00:00:01 mcast address to master interface
    Dec 31 16:00:08 FreeWilly user.info kernel: vlan1: dev_set_allmulti(master, 1)
    Dec 31 16:00:09 FreeWilly user.info kernel: vlan1: dev_set_promiscuity(master, 1)
    Dec 31 16:00:09 FreeWilly user.info kernel: device eth0 entered promiscuous mode
    Dec 31 16:00:09 FreeWilly user.info kernel: device vlan1 entered promiscuous mode
    Dec 31 16:00:09 FreeWilly user.info kernel: device eth1 entered promiscuous mode
    Dec 31 16:00:09 FreeWilly user.info kernel: br0: port 2(eth1) entering forwarding state
    Dec 31 16:00:09 FreeWilly user.info kernel: br0: port 1(vlan1) entering forwarding state
    Dec 31 16:00:09 FreeWilly user.warn kernel: vlan2: Setting MAC address to  bc ae c5 e8 14 a1.
    Dec 31 16:00:09 FreeWilly user.debug kernel: vlan2: add 33:33:00:00:00:01 mcast address to master interface
    Dec 31 16:00:09 FreeWilly user.debug kernel: vlan2: add 01:00:5e:00:00:01 mcast address to master interface
    
    So internally within the hardware and firmware, everything consistently identifies these ports as "1" and "2" (in spite of ASUS' PDF manuals).

    But build 108 still has the ports labelled as "LAN 3" and LAN 4". And "Invert Ports Order" does not fix this.
     
  18. M_ars

    M_ars LI Guru Member

    Hi shibby,

    i think you forgot to change two files for QoS Prios (with up to 10 classes) --> see pkt_sched.h in linux-2.6 and linux folder

    #define TC_HTB_NUMPRIO 11 (right now you have 10)
    and
    #define TC_CBQ_MAXPRIO 11 (right now you have 10)

    ==> see commit "fix QOS prios" from Toastman: 636e61fffa8c70a92082d7b0474b488da6d441db

    http://repo.or.cz/w/tomato.git/commit/636e61fffa8c70a92082d7b0474b488da6d441db

    ==> Can you also add commit "IPTraffic bugfix" from Toastman (thx Merlin): 3f00df5ef808df3581df2c3084bff7d4c7c2246d

    http://repo.or.cz/w/tomato.git/commit/3f00df5ef808df3581df2c3084bff7d4c7c2246d



    One request for Asus RT-N12 --> Can you please add the following code in "advanced-vlan.asp"?

    Code:
    case '0x04cd': //RT-N12
    case '0x054d': //RT-N12B1
    if ((nvram['boardrev'] == '0x1201') || (nvram['boardrev'] == '0x1101')) {
    COL_P0N = '3';
    COL_P1N = '2';
    COL_P2N = '1';
    COL_P3N = '0';
    COL_P4N = '4';
    break;
    }
    Thank you :)
     
    shibby20 likes this.
  19. vlads

    vlads Serious Server Member

    Folks, I'm sure I'll get flamed for this, but don't you think the Tomato project is fast becoming a runaway train?

    What has for a long time been focused on improving the functionality of a decent retail product, is turning into more of a fashion product or a toy.

    Every day I see requests for blinking port lights, user defined port lables and for themes and other junk. In my opinion a router is meant to do just that: route, qos, trafic shaping, VPN etc.

    Should the device have enough horsepower it could be used as a NAS or soft PBX or (fill in the purpose) but that should be secondary and handled via optware/entware. All these extra gimmicks we keep asking to be added in the main image will r the images bigger and bigger, will fill up the NVRAM and ultimately impact stability.
     
  20. Elfew

    Elfew Addicted to LI Member

    I dont agree with your opinion... Yes, Tomato FW is mainly focused on stability and performance.... there are many builds, you can choise one with specific features... Nowadays there are modern routers with 64kB NVRAM and 256MB ram for very low price... if you have the old one, just use another build without extra features...

    There are only 3-4 active developers, I think there should be a list or something like that and every developer should focus on one new feature/thing... There is new openvpn version, but nobody create a commit for tomato; some modules in tomato are outdated, we should find a solution - every developer should contribute to the main Tomato FW branch which would be the official and tested version. This version would be maintained by themselves.

    What is your opinion? I think Tomato needs a person who can lead this project!
     
    philess likes this.
  21. RMerlin

    RMerlin Network Guru Member

    That's the beauty of open source: you have plenty of variants to chose from. Some, like Toastman's builds, are just what you are looking for. Other variants will appeal more to people who like to tinker, as a hobby. I think it's great that, as a Tomato user, you get to chose between the barebone version, or the kitchen-sink version.
     
    SNR likes this.
  22. RMerlin

    RMerlin Network Guru Member

    I think Tomato shouldn't be looked at as "a project", but as a gathering of forked projects, sharing a common base, but offering alternatives.

    Kinda like Linux (the whole OS, not just the kernel). You have distros for all tastes, without the need for a "project lead". Linus only handles the Kernel, and doesn't tell each distro which desktop to use (or anything related to userspace in fact). And in the end, there's something for everyone.
     
  23. Elfew

    Elfew Addicted to LI Member

    I agree. But there should be a list of request or something like that... Because sometimes 2 persons work on the same thing, it is waste of time, dont you think RMerlin?

    for example - there is a new updated version of something (busybox or something else).. 3 developers started working on it, every on your own... why they dont distrubute this work - one will apply new version to the tomato fw and others just will take this commit... so others developers would have more time for something else - update next outdated version, fix some bugs etc.

    I hope you understand me :) English is not my native language ;)
     
  24. eahm

    eahm LI Guru Member

    As a project yes but like Linux no. The Linux community is big and awful, there are 746 "different" distro and it's just a big confusing mess. People create a fork, changing just the wallpaper and they call that a "new distro", awful awful awful!

    I remember when Linux was a project and there were only 4-5 distro to choose and it was amazing and the possibility to destroy the big MS conpetition was very high. Now it's just a big joke, good only for servers.

    I've been using Linux for 16 years and still don't know which one to use, forks and forks. I'll stick with Windows.
     
  25. RMerlin

    RMerlin Network Guru Member

    I understand your point about coordinating developments to avoid overlap or wasted resources. But I don't think this has been an issue so far. The forums are already a good place to start if you want to have an idea of what is being worked on by others. I don't think Tomato is at a stage where it would require such controlled coordination.
     
    Elfew likes this.
  26. dr.gazza

    dr.gazza Networkin' Nut Member

    Hi,
    I'm here to ask for help!

    I use a E4200 with internet key for the internet;

    The internet key has a slot for microSD, in which I have inserted a microSD card to be mounted on tomato shibby
    to store the log file, but unfortunately I can not do the mount;

    Can anyone help me?

    Thanks in advance

    Router: E4200 with Tomato Shibby 1.28.0000 MIPSR2-102 K26 USB AIO
     
  27. SNR

    SNR Networkin' Nut Member

    @vlads, I don't really need the eyecandy myself, but in the case of Ethernet Ports State, I was just reporting a bug that I found. I can certainly live without that feature myself.

    FWIW, I did revert to build 106 temporarily and noticed that in that version, Ethernet Ports State has a single "LAN" label above all four LAN ports. This explains why I didn't notice the issue when it was first introduced. And ASUS' stock firmware also exhibits the same bug...

    @RMerlin, I just switched over to your Asuswrt-Merlin mod a few days ago after realizing that I currently don't need most of the features that Tomato has anyway. The Swiss Army knife approach is cool, but as General Motors used to say, "Fewer moving parts means less to go wrong."

    In any event I will probably continue to flash back and forth between your mod and Shibby's (and Toastman's) just to see the current state of the art. At some point it becomes all about the journey, not the destination. ;)

    I see that you have a bug report in your GitHub relating to this issue:

    Switch ports are in the reversed order
    https://github.com/RMerl/asuswrt-merlin/issues/262

    Looking forward to your next build to see how you did with this. Thanks!
     
  28. RMerlin

    RMerlin Network Guru Member

    I simply establish the port order as set by the firmware, and label using that order:

    https://github.com/RMerl/asuswrt-merlin/commit/64b062d7fb525a5299e35aa6e9cd3fb7a0c00362

    I should be receiving an RT-N16 in a week or two, will allow me to confirm that it works properly.
     
    SNR likes this.
  29. SNR

    SNR Networkin' Nut Member

    That's great news, hopefully you can start officially supporting the RT-N16?

    BTW, I just sent you a small token of appreciation. Not much but it should help pay a little towards that router. Thanks again!

    Edit: I just noticed that I misspelled this router's model name all over the last two pages of posts. It's RT-N16, not RT-16N. I'm not going to bother going back and cleaning up every post. I'll just leave the errors as a testament to my tendency towards "tactile dyslexia" (I read much better than I type).

    @Elfew, check your signature... ;)
     
  30. Nitro

    Nitro Networkin' Nut Member

  31. RMerlin

    RMerlin Network Guru Member

    I already know they are backward. But I don't want to implement a manual switch like shibby20 did, and rely instead on the firmware to tell me the correct order. That way, if Asus were to ever release a new version that had the ports ordered in a different order, it would automatically be picked up by the FW, without having an additional, confusing webui option.
     
    pharma likes this.
  32. leandroong

    leandroong Addicted to LI Member

    For info: Transmission modder has just fixed magnet issue on revision#14061.
     
  33. Huey

    Huey Networkin' Nut Member

    Hello all,

    Sorry to bother you all with a small question about NFS:
    I want to share my USB disk formatted as ext3 over NFS. Actually I think have been able to mount to it using the NFS client in Windows 7. (The /etc/export file is looking ok, on windows I have done; "mount \\192.168.1.1\USB X:" where the mount is located /tmp/mnt/USB).
    I was thinking of disabling SAMBA file sharing now, but then my mount fails. Is it supposed to also be left on, or is it actually still mounted over samba and is my NFS configuration not ok?

    Thanks for your time,

    Huey.

    Edit: my apologies. took a pause and did some proper retesting. Disabled the SAMBA sharing and used a linux box first to try to mount the NFS share. This was succesfull...
    Then used "showmount -e 192.168.1.1" on the windows box, and that was also sucessfull. Mount still gave me an error 53: seems that you also need to set the order windows uses mount options. Going to try that now.

    Edit2: Working now. Indeed need to set NFS on top of provider order. Also read that one must add AnonymousUid and Anonymousgid to registry. I do have the feeling that it only started working after I made 1 succesfull connect with the opensource Nekodrive NFS client first...

    Edit3: NFS is slower from the windows box (about 5700 kbytes/s) compared to samba (7900 kbytes/s) to ext3 formatted USB drive
     
  34. shibby20

    shibby20 Network Guru Member

    another really nice addition to Tomato: ipset module.

    I successful included this package into Tomato. It will be included in next release :)

    Many thanks RMerlin :)
     
    Fredrik and kthaddock like this.
  35. FireWire

    FireWire Serious Server Member

    Hi. Hope someone can help me out. My router is ASUS RT-N66U. I recently started to have problems with the wireless connection to my media player(WDTV Streaming Media Player). I never had any issues before. But after I upgraded to 64K version there seem to be issues that I didn´t have before. I am running Shibbys firmware 108AIO 64K version. I also tried the 107 version, but same issues there. I will attach an image. I guess it will seem quite funny to complain about the values, when they are too good =). I always had negative RSSI-values and quality about 30-60. But suddenly I started have positive RSSI-values and quality that went over 100!?! Should that even be possible, shouldn´t 100 be max? The thing is that with these values everything I play on the media player is choppy.
    I tried changing transmit power to see if it would make any difference, but it doesn´t change anything. Tried to set a value of 0(hardware default), 50mW and 100mW. After testing I went back to 50 again.
    I really hope someone can give me a clue on what is going on here.

    [​IMG]
     
  36. kthaddock

    kthaddock Network Guru Member

    @FireWire
    Your quality should be, 60 that value takes som time to display properly. After upgrade of program did you do a NVRAM reset through web gui drop list? I think that comes from old config isn't wiped out properly.
    Reset and reconfigure by hand, do not use saved config file.

    kthaddock
     
  37. FireWire

    FireWire Serious Server Member

    Ahh, that might be the issue then. Did NVRAM reset, but applied a saved config file from 108 version(32K NVRAM version).
    Have something to do then... Will try it as soon as I have the time.
    Big thanks for the fast answer and suggestion kthaddock, I appreciate it.

    Best regards,
    FireWire
     
  38. FireWire

    FireWire Serious Server Member

    Ok, have done a NVRAM reset and configured everything from scratch. But, same issue still occur. The RSSI values are positive and the quality is above 100. Any suggestions on what´s going on here?
     
  39. callous

    callous Network Guru Member

    I was using the Openvpn client in Shibby's 106 and was wondering if it meant my openvpn session died and the openvpn client reconnected shortly after.

    unknown daemon.err openvpn[19080]: event_wait : Interrupted system call (code=4)

    Below is the log:

    Code:
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: OpenVPN STATISTICS
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: Updated,Tue Apr 14 12:42:58 2013
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: TUN/TAP read bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: TUN/TAP write bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: TCP/UDP read bytes,3928
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: TCP/UDP write bytes,1517
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: Auth read bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: pre-compress bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: post-compress bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: pre-decompress bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: post-decompress bytes,0
    Apr 14 12:42:58 unknown daemon.notice openvpn[19080]: END
    [B]Apr 14 12:57:59 unknown daemon.err openvpn[19080]: event_wait : Interrupted system call (code=4)[/B]
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: OpenVPN STATISTICS
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: Updated,Tue Apr 14 12:57:59 2013
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: TUN/TAP read bytes,25173259
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: TUN/TAP write bytes,147566751
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: TCP/UDP read bytes,152561067
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: TCP/UDP write bytes,28778689
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: Auth read bytes,147566783
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: pre-compress bytes,791043
    Apr 14 12:57:59 unknown daemon.notice openvpn[19080]: post-compress bytes,805092
    Apr 14 13:00:01 unknown syslog.info root: -- MARK --
    Apr 14 14:00:02 unknown syslog.info root: -- MARK --
    Apr 14 15:00:01 unknown syslog.info root: -- MARK --
    Apr 14 16:00:01 unknown syslog.info root: -- MARK --
    [B]Apr 14 16:41:26 unknown daemon.err openvpn[19080]: event_wait : Interrupted system call (code=4)[/B]
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: OpenVPN STATISTICS
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: Updated,Tue Apr 14 16:41:26 2013
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: TUN/TAP read bytes,445210822
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: TUN/TAP write bytes,2415037508
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: TCP/UDP read bytes,2503303913
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: TCP/UDP write bytes,510153939
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: Auth read bytes,2415037540
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: pre-compress bytes,14671660
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: post-compress bytes,14943294
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: pre-decompress bytes,543
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: post-decompress bytes,568
    Apr 14 16:41:26 unknown daemon.notice openvpn[19080]: END
    [B]Apr 14 16:41:44 unknown daemon.err openvpn[19080]: event_wait : Interrupted system call (code=4)[/B]
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: OpenVPN STATISTICS
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: Updated,Tue Apr 14 16:41:44 2013
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: TUN/TAP read bytes,445479262
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: TUN/TAP write bytes,2415211139
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: TCP/UDP read bytes,2503508612
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: TCP/UDP write bytes,510450603
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: Auth read bytes,2415211171
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: pre-compress bytes,14671660
    Apr 14 16:41:44 unknown daemon.notice openvpn[19080]: post-compress bytes,14943294
    Apr 14 17:00:01 unknown syslog.info root: -- MARK --
     
  40. koitsu

    koitsu Network Guru Member

    The log indicates your VPN session is going down (either due to a timeout or due to network anomalies); errno 4 == Interrupted system call, and the system call being interrupted is probably read() or write().

    You're going to need to increase the logging on OpenVPN to get even more detail, or you're going to need to start monitoring your network connection to the VPN server 24x7x365 using tools like mtr, WinMTR, PingPlotter, or other such things. Alternately you may be having issues with your ISP or your physical line to your ISP, in which case OpenVPN isn't the only thing being impacted when this happens.
     
  41. callous

    callous Network Guru Member

    Ok thanks! I will have to troubleshoot this in addition to weird router reboots in the past few weeks while running openvpn client on the router
     
  42. shibby20

    shibby20 Network Guru Member

    i had this error when ovpn server was in very old version.
     
  43. callous

    callous Network Guru Member

    Thanks!

    I know my vpn connection continues despite this error. My downloads certainly didnt slow down that I could see. Maybe I shouldnt worry about it.
     
  44. koitsu

    koitsu Network Guru Member

    Figuring this out is not easy, because it could be any syscall that's returning EINTR. The only way to find out would be to run OpenVPN under strace and log all the crap to a file somewhere for review later. (It would be enormous, by the way)

    The EINTR situation happens then OpenVPN immediately begins spitting out a bunch of "STATISTICS". This happens every time for every one of your log entries (look very closely at the timestamps).

    I can't find details about what this "OpenVPN STATISTICS" thing is. All I can find on the net is this thing, which looks like some kind of "plugin" that doesn't really tell me jack squat. For all I know it might be trying to read from an fd in a non-blocking state with a timeout and then get handed EINTR once the timeout value is reached; the string "event_wait" implies a timeout of some kind, but again, I don't know what. I'm having to make guesses since I have no strace data or nor have I reviewed the code (which is not necessarily going to shed light on it anyway -- most of this stuff is highly abstracted and is the biggest rabbit hole you've ever seen).
     
  45. callous

    callous Network Guru Member

    As it turns out, simply having the openvpn client running in the router, then after a few dozen minutes going into the Openvpn Client>status page will show a event_wait : Interrupted system call (code=4) in the log.

    The connection isnt interupted. However, that is only one of the ways to get a code=4. The other instances of this happening I will have to enable logging I suppose.
     
  46. RMerlin

    RMerlin Network Guru Member

    When you access the VPN Status page, it signals the VPN server/client with SIGUSR2. That causes it to dump its statistics both in syslog and in a text file, which then gets loaded by the httpd daemon, parsed, and returned to the webui client.
     
  47. koitsu

    koitsu Network Guru Member

    Thank you -- this is exactly what I needed to know. That absolutely explains what's going on. The syscall being executed by openvpn at the time (likely guesses are select(), poll(), read(), or write()) are interrupted as a result of the process receiving SIGUSR2.

    So no, in this case, the messages shown are not indications of a network issue or OpenVPN going down/having issues -- it's a result of the methodology/model used to pull statistics from OpenVPN.

    It's very possible to inhibit the logging line in question for EINTR when SIGUSR2 is sent (sigsetmask(), sigblock(), and similar come to mind). Please contact whatever author wrote said code and inform him/her of the situation; this is something that should be fixed upstream, not within TomatoUSB.
     
  48. cliffi

    cliffi Network Guru Member

    iam using Shibbys Tomato Firmware 1.28.0000 MIPSR2-108 K26 USB VPN on my Asus RT-N16 and getting alot of this warnings in my logfile, anything to worry about?

    Apr 17 23:46:31 unknown daemon.warn miniupnpd[17532]: upnp_event_process_notify: connect failed: Connection reset by peer
    Apr 17 23:46:45 unknown daemon.warn miniupnpd[17532]: upnp_event_process_notify: connect failed: Connection reset by peer
    Apr 17 23:46:59 unknown daemon.warn miniupnpd[17532]: upnp_event_process_notify: connect failed: Connection reset by peer
     
  49. RMerlin

    RMerlin Network Guru Member

    The newer miniupnpd added support for UPNP Events. These can generate a lot of similar warnings, and can safely be ignored.

    On my end I have increased the syslog level of these so they won't be logged at the normal loglevel.
     
    koitsu likes this.
  50. canuckerfan

    canuckerfan Serious Server Member

    this might be a little off-topic but how are you guys generating keys using openvpn? I'm always getting openssl errors?
     
  51. Mangix

    Mangix Networkin' Nut Member

    https://github.com/OpenVPN/easy-rsa
     
  52. cliffi

    cliffi Network Guru Member


    thanks. ok, good idea
     
  53. callous

    callous Network Guru Member

    I was just looking through my log and found A LOT of the following entries. Using Shibby 106 (because 107 and 108 5ghz wireless ceases to work after a few days)

    What does the following error that keeps showing up over 100 entries in an hour mean?

    unknown user.warn kernel: wlc_print_ampdu_txstatus: txstatus 0x1063
     
  54. koitsu

    koitsu Network Guru Member

    Google is your friend.

    These are messages from the binary-blob (meaning we do not have source for it) wireless driver called "wl". Nothing can be done about the messages because the source to the wireless driver is unavailable.

    You could try asking Broadcom directly. *laughs*

    Regardless, these messages have nothing to do with what this thread is about (Tomato RAF with included BitTorrent Client).
     
  55. alpovs

    alpovs Reformed Router Member

  56. koitsu

    koitsu Network Guru Member

    I believe you are conflating two unrelated issues. Your problem in that thread is that wireless devices on your network don't disappear after many hours (or days) of them physically being out of range or powered off; FireWire is talking about something different.
     
  57. alpovs

    alpovs Reformed Router Member

    No, my problem is that new devices (newly turned on) cannot connect to wireless. I wouldn't care if devices didn't disappear. It was just an additional observation. FireWire and I have a common observation about the senseless levels. I think it is a part of the same issue. FireWire mentioned that his observations began when he switched to 64k builds. I have always used 64k builds and I always had this issue. Maybe it has something to do with type of builds? I suggest to take this discussion to that new thread: http://www.linksysinfo.org/index.ph...eless-behavior-on-rt-n66u-possible-bug.68438/
     
  58. koitsu

    koitsu Network Guru Member

    The NVRAM content size (64KB vs. 60KB vs. 32KB) has no bearing: period.

    Your problem, by the way, seems to be (as of this writing -- it's the only thing we can go off of) related to the wireless driver becoming "wedged" and requires jostling (by either rebooting the router or using the Noise Floor check, both which reset the wireless driver + partially reinitialise the wireless chip). The most common cause for this issue is excessive noise within specific frequency bands, which can include things like other wireless APs, but more commonly things like baby monitors, microwaves, USB 3.0 (yes really, Intel did a study on it), and tons of other things I can't remember. It's all stuff you can't see with a "Scan" either -- the only stuff that sees it is a hardware wireless protocol analyser.

    Now about the "Quality" field -- AND ONLY ABOUT THAT FIELD -- yes, that is a "made up" value by Tomato/TomatoUSB, believe it or not. It's based on the following formula (comes from the source code in release/src-rt/router/www/status-devices.asp):

    quality = MAX(rssi - noisefloor, 0);

    The Javascript function MAX() is defined in tomato.js and is a wonderful example of someone who is a complete jerk (forcing the reader to know Javascript operator order/precedence else misunderstand what it does -- because using parenthesis to make it easier to understand is just too hard...):

    Code:
    function MAX(a, b)
    {
            return a > b ? a : b;
    }
    
    This equates to:

    Code:
    if a > b
      return a
    else
      return b
    
    And should really be rewritten into return (a > b) ? a : b; just for legibility.

    So basically if rssi-noisefloor > 0, then return rssi-noisefloor, otherwise return 0 -- and that's the quality.

    If you haven't calculated your noise floor, then this number may appear "different" than if you have calculated your noise floor. The noise floor value, once obtained by the wireless chip itself, is stored in NVRAM variables ending in _tnoise (i.e. nvram show | grep _tnoise). Please do not get into a discussion about the wl_xxx vs. wl0_xxx variables; I don't have the time/interest/etc. to explain that right now.

    There is a lot more logic applied here in the code, and due to its lack of comments, is a bit hard to follow. For example I see some code where non-clients end up with a noise value of -99 if the noise value obtained is >= 0 or < -100 (this is hard-coded in get_wlnoise()).

    You have to understand something -- digging through the Tomato source is like going down one of the biggest rabbit holes on earth, with no comments, no help, no nothing. It requires many, MANY hours of reverse-engineering. I hate using this stupid excuse (I really do), but this is a best-effort project, and not all of us can sit around banging on hours of questionable code. Money is not an incentive either -- personal interest is. That's just how the project works.
     
    jerrm and Victek like this.
  59. kthaddock

    kthaddock Network Guru Member

    @Shibby

    I have played with UpnP and custom config. There is two thing I aske to be fixed.
    1: Remove upnp check when no bridge is set, so you can use custom config even if you have only one lan bridge configured. ( see pic )
    2: Length of "upnp_custom=" seem to short, eg when you have two deny rules for two bridges only one fit.
    But they are in config file both of them. I'm not sure "deny" rules working as they should either.
    This: nvram export --set | grep upnp
    Best regards
    kthaddock
     

    Attached Files:

  60. JakeS

    JakeS Reformed Router Member

    Hi, firstly thanks to Shibby for a great FW!

    I have 1.28.0000 MIPSR2-105 K26 USB BT-VPN installed on my WNR3500Lv1 and overall it's fantastic, but one thing seems to be amiss with the authentication of Samba shares.

    I have a USB flash drive mounted and it shows at the bottom of the USB support page. On the Samba sharing page I have "Yes, no authentication" selected. If I try to browse the USB stick from a local Windows PC I can do so initially, after I have set "Yes, no authentication". However, if I reboot the router, I can no longer browse the USB stick - I am denied access and it asks for a username and password. This is even though "yes, no authentication" is still showing in the Enable File Sharing dropdown. I found the way to turn off authentication is to select another option in the dropdown, then reselect "yes, no authentication" and save settings. Then I can browse again. But as soon as I reboot the router the setting is lost. (This happens regardless of whether Master Browser is ticked or not.)

    I tried enabling authentication and setting a username and password, but then it will only let me login from one local PC at a time. If I go from my desktop to my laptop, for example, I get a message saying the USB drive is already in use.

    If there's a solution it would make using the USB stick as NAS much easier.
     
  61. leandroong

    leandroong Addicted to LI Member

    use CF card as NAS, very reliable.
     
  62. big g

    big g Reformed Router Member

    I had very similar issues with my e3000 on both 106 and 108. I actually had to upgrade to 108 because my hard drive accidently got unplugged and I couldn't access it over the network after trying to switch authentication methods. Even after updating to 108, I had to reset the NVRAM a couple times and tinker with the authentication modes in order to get it working (especially with Transmission). Hopefully this will become more stable in the upcoming builds.

    I'm glad I finally got it working, but I'm worried in case something happens in the future.
     
  63. tenskwatawa

    tenskwatawa Networkin' Nut Member

    @FireWire
    Try to configure your wifi network with Channel Width on 20 MHz. It works without this issue.
     
  64. GhaladReam

    GhaladReam Network Guru Member

    hey shibby,

    It's been a while since your last update. Just wondering what we might be able to expect for build 109 and beyond?
     
  65. kyrios

    kyrios Serious Server Member

    - USB Modeswitch 1.25
    - Delete VLAN2 problem, fixed
    - DNSMasq 2.67 with all patches
    - Temp Info
    - Deletion Restriction Access problem using Chrome and Firefox; fixed
    - 2TB HD support (OOM?)

    I guess those are you can expect from v109, right Shibby?
     
  66. anpak

    anpak Networkin' Nut Member

    Hi guys!!!
    I've a question for Shibby...is possible insert a L2TP/IPSec Server in your firmware?:)
    thanks a lot in advance!!!;)
     
    Elfew likes this.
  67. zorkmta

    zorkmta LI Guru Member

    Are there any ETA released day for v 109?

    Enviado desde mi C6603
     
  68. JugsteR

    JugsteR Serious Server Member

    Yes there is. However not communicated publicly.

    Signature? Signature!
     
  69. shibby20

    shibby20 Network Guru Member

    @zorkmta - for a day, maybe two. K24 and K26 are ready. RT-N is compiling at the moment.

    @anpak - I don`t know L2TP and IPsec server configuration well i`m not a good person to do this. For IPsec you have all needed modules in my tomato (AIO and Mega, for others buildtypes modules for ipsec are in extras). Then install entware and install strongswan, configure and run.

    @kyrios - changelog will be different :)
     
    Fredrik, Elfew, zorkmta and 1 other person like this.
  70. Frequenzy

    Frequenzy Networkin' Nut Member

    cant wait for 109 :)
     
  71. JugsteR

    JugsteR Serious Server Member

    I can :)

    Signature? Signature!
     
  72. hawkmat

    hawkmat Networkin' Nut Member

    The wireless signal strength on the Device List page does not work correctly. No adjustments you make will fix this problem. The wireless drivers have been updated a few times lately with AsusWRT 5.100.138.20 drivers. This rendered the RSSI reading inoperable on the Device List page. It only works reliably with the old 5.100.138.9 driver in an older version of Tomato. I recommend that you set your output power to 0 for both bands and use the Device List to only see who is connected and ignore the signal strength.
     
  73. shibby20

    shibby20 Network Guru Member

    [RELEASE] v109

    Please read changelog first.

    btw 3500Lv2 builds are compiling right now. Will be soon as possible.
     
    M_ars, jerrm, Elfew and 1 other person like this.
  74. Badders44

    Badders44 LI Guru Member

    Nice one Shibby. Loaded and running. No issues so far. :) Certainly no boot delay ;)
     
  75. lefty

    lefty Networkin' Nut Member

    The E3000 109 RT Mega build is much to large to fit on an E3000. The image is 8.6MB and the E3000 only has an 8MB flash.
     
  76. leandroong

    leandroong Addicted to LI Member

    ZTE ZXV10 H618B
    No issue also. Same boot time. What is extended MOTD?
     
  77. shibby20

    shibby20 Network Guru Member

    MOTD - Message Of The Day

    Just log in via SSH and you will see ;)
     
  78. kthaddock

    kthaddock Network Guru Member

    Thanks for new build, I'm gone test tonight.
    Does this build include: "string" module?

    kthaddock
     
  79. shibby20

    shibby20 Network Guru Member

    no, it doesn`t.
     
  80. Badders44

    Badders44 LI Guru Member

    I really like the SSH MOTD info but is the "used" memory correct?

    From overview:

    Total / Free Memory123.87 MB / 107.98 MB (87.17%)

    From MOTD:

    Mem usage: 90.3% (used 111.90 of 123.87 MB)
     
  81. shibby20

    shibby20 Network Guru Member

    because you have enabled "Count cache memory and buffers as free memory" in Administration -> Debugging. Disable this option, save and re-check once again.
     
  82. Badders44

    Badders44 LI Guru Member

    Jeez - more haste, less speed..apologies

    From overview:

    Total / Free Memory123.87 MB / 11.67 MB (9.42%)

    From MOTD:

    Mem usage: 90.7% (used 112.37 of 123.87 MB)
     
  83. kthaddock

    kthaddock Network Guru Member

     
  84. soooooil

    soooooil Reformed Router Member

    Hello Shibby, thanks for your work, your builds are the best! But in all last releases there is a problem with 4 MiB routers (I have e1200v1), firmware is too big to fit (as I understand, maximum file size that will fit is 3866624 bytes). Can something be done, like removing few themes or unessential features, or does this mean that there will be no more releases to fit in 4 MiB?
     
  85. sanmao

    sanmao Reformed Router Member

    Thank you for this great firmware, Shibby!
    is v109 drop max build?
    I use tomato-K26-1.28.RT-N5x-MIPSR2-108-Max.trx, on my Asus rt-n12B1.But i can't find max build in v109.
    Thanks.
     
  86. tomatosoup

    tomatosoup Serious Server Member

    Thanks for 109 shibby, works great on my E4200.

    I also flashed it (tomato-ND-1.28.5x-109-VPN.trx) on an WRT150N (I know, it's not officially supported - recognized as WRT300N) - working also fine there, except Ethernet Ports State does not work there - all ports show just 'unplugged'. Don't know if it was working with older builds, I ran 103 or 104 for a long time there. Does not really matter to me, just wanted to report...
     
  87. shibby20

    shibby20 Network Guru Member

    propably switch is not supported by robocfg. can you run command robocfg show and let me know what is a result.
     
  88. tomatosoup

    tomatosoup Serious Server Member

    Here you go:

    Code:
    Switch: enabled
    Port 0:  DOWN enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    Port 1:  DOWN enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    Port 2:  DOWN enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    Port 3:  DOWN enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    Port 4:  DOWN enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    Port 5:  100FD enabled stp: none vlan: 1 mac: 00:00:00:00:00:00
    VLANs: BCM5325/535x disabled mac_check mac_hash
    
     
  89. Kibe

    Kibe LI Guru Member

    Hello everyone!

    I'd like to redirect all traffic from a specific IP address range in my network to TOR, instead of just the traffic with destination port 80 as described in the GUI.

    I'm using Tomato Firmware 1.28.0000 MIPSR2-108 K26 USB AIO-64K on a Asus RT-N66U.

    Is this possible with maybe some custom configuration on that field?

    Thanks for checking!
     
  90. shibby20

    shibby20 Network Guru Member

    @tomatosoup - as i said, switch in your router is not supported by robocfg and all ports are detect as DOWN. We have to wait for new version of robocfg. Sorry.
     
  91. RMerlin

    RMerlin Network Guru Member

    Note that robocfg is deprecated now, and will probably never be updated. People have rewritten it as a kernel module, it is being used by OpenWRT and DD-WRT now. These support more switches, but they don't work the same way. Search for switch-robo.
     
  92. Bird333

    Bird333 Network Guru Member

    I know that I could use vlans on my wrtsl54gs under openwrt. I got Eko to implement it on DD-WRT. He had to do something with 'boardflags' and my cfe if I remember correctly. Of course Shibby doesn't support vlans on this router but I know it is possible.
     
  93. tomatosoup

    tomatosoup Serious Server Member

    @shibby: Thanks for investigating - and don't worry, this feature is not essential to me, I just wanted to report that it's not working on this device.
     
  94. GhaladReam

    GhaladReam Network Guru Member

    @shibby - Thanks for the new build (109), appears to be working well on all my units :) I like the animated ports state.
     
  95. uboot

    uboot Serious Server Member

    I've a little request: when I shutdown my router I would see all connected hdd go to sleep at once. Maybe halt script may call scsi-stop for all drives. Thanks.
     
  96. koji

    koji Networkin' Nut Member

    Anyone of you have issues with IPv6 or it's just me? I have a dual-stack DHCPv6 with Prefix Delegation from my ISP and after flashing 109-AIO, IPv6 is not working anymore. Should I revert back to 108-AIO?
     
  97. JugsteR

    JugsteR Serious Server Member

    Did 108 work? Then yes. Have you tried clearing nvram? Then no, try that first.

    Signature? Signature!
     
  98. vlads

    vlads Serious Server Member

    @shibby - both MIPS1 and 2 109K26 MiniVPN builds are significantly larger than the 4MB routers they were targeted for.

    Is there any plan to shrink them down further by removing additional features or should those images be deprecated?
     
    soooooil likes this.
  99. koji

    koji Networkin' Nut Member

    @JugsteR,
    I already tried 3 times, cleared NVRAM...
    It's strange anyway, because when I ping a IPv6 adress from router system everything is working flawless, but when I try to acces that adress from my PC I receive the classic "Could not locate remote server".
    I don't have same issue with 108-AIO so I'll stick with it...maybe I have more luck next time, at 110-AIO.
     
  100. fubdap

    fubdap Addicted to LI Member

    Did you see this on a different post. Not sure if it's the same problem.
    http://www.linksysinfo.org/index.php?threads/shibby-109-ipv6-tunnel-broken.68559/#post-228033


     

Share This Page