1. Forum
  2. FidoNet
  3. SYNC SYSOPS

  • Disconnected users cannot reconnect occasionally.

    From Nigel Reed@1:103/705 to GitLab issue in main/sbbs on Thu Feb 8 22:37:51 2024
    open https://gitlab.synchro.net/main/sbbs/-/issues/713

    I've had a couple of users mention this so thought I'd open an issue. The latest one said that their wifi disconnected, which dropped their BBS connection. When he tried to relog in the BBS said he was already connected. It seems the BBS didn't timeout his connection for quite a while.

    This is probably going to be one of those random hard to pin down issues, but I thought I'd at least log it.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Thu Feb 8 22:50:21 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4785

    Using what protocol? If telnet, you can detect disconnected users more quickly by enabling the "Send Go-Aheads" feature.

    Otherwise, you can TCP KEEPALIVE in ctrl/sockopts.ini to more quickly detect disconnected users. Or try tuning some of the TIMEO values in sockopts.ini. Your network stack may have configuration parameters for TCP session timeouts and liveliness checks that are via sock opts as well.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Nigel Reed@1:103/705 to GitLab note in main/sbbs on Fri Feb 9 23:42:04 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4788

    My user said, and I quote:

    ```
    One problem I've been having is that when I go to certain screens, the BBS
    seem to lock up. After a minute or so, I'll give up and disconnect. The problm then is, when I go to reconnect, it won't let me log in for hours because it believes I'm already logged in.
    ```

    It wouldn't be so bad if it was 10 minute, 30 minutes or whatever, but I don't even see Linux having a socket timeout of hours. It's like it's not detecting he logged out at all. Maybe he only got back in because I did a recompile and shutdown and restarted.

    I've asked him to notify me if it happens then I can check the state of the socket. When I have charged my laptop I can give it a try and see if I can replicate the issue by dropping wifi.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 00:55:24 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4789

    Set SCFG->Servers->Terminal Server->Max User Inactivity to some value?
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Keyop Troy@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 13:44:42 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4794

    I have the same problem and end up giving regular users the G exemption so that they can login to multiple nodes at the same time.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Keyop Troy@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 13:49:35 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4795

    I’m going to try that on mine.

    Would 10 mins be a suitable value? Why is it 1 hour for new users?

    Thanks DM.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Nigel Reed@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 14:26:14 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4797

    I set mine to 10 minutes. There's really no reason for a user to idle that long anyway.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Keyop Troy@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 14:30:45 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4798

    Nelgin: what have you changed to 10 mins? The max user inactivity or max new user inactivity?
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Nigel Reed@1:103/705 to GitLab note in main/sbbs on Sat Feb 10 14:39:45 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_4799

    I have max login inactivity and max user inactivity set to 10 minutes and new user inactivity is 30. I'll probably change that to 10 also.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Wed Aug 7 23:33:13 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_5471

    Seems we can close this issue now then (?).
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Nigel Reed@1:103/705 to GitLab note in main/sbbs on Thu Aug 8 00:50:46 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_5493

    I'm not sure the underlying issue ever got resolved in that a user can be disconnected but the socket stays open until it times out but very difficult to replicate.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Thu Aug 8 01:28:13 2024
    https://gitlab.synchro.net/main/sbbs/-/issues/713#note_5499

    Most likely, this: https://superuser.com/questions/298919/what-is-tcp-half-open-connection-and-tcp-half-closed-connection

    The socket inactivity detection feature can be used to successfully detect this condition.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)