| Both sides previous revisionPrevious revisionNext revision | Previous revision |
| faq:tcpip [2020/04/17 20:19] – [Bind] expand on the rebind issue. digital man | faq:tcpip [2020/06/01 21:33] (current) – [Rebind] TIME WAIT or CLOSE WAIT - both are indications of the same problem digital man |
|---|
| This error upon binding usually means that you're running Synchronet as non-privileged user account (e.g. not 'root') and the operating system you're running does not allow processes run under non-privileged user accounts to bind to low (TCP or UDP) port numbers (usually less than 1024). You can either use higher TCP port numbers in your configuration or have Synchronet switch to a non-privileged user *after* binding the TCP ports (see [[config:nix]] for details), see also: [[howto:Linux non-root]]. | This error upon binding usually means that you're running Synchronet as non-privileged user account (e.g. not 'root') and the operating system you're running does not allow processes run under non-privileged user accounts to bind to low (TCP or UDP) port numbers (usually less than 1024). You can either use higher TCP port numbers in your configuration or have Synchronet switch to a non-privileged user *after* binding the TCP ports (see [[config:nix]] for details), see also: [[howto:Linux non-root]]. |
| |
| | ==== Rebind ==== |
| :!: **Answer:**\\ | :!: **Answer:**\\ |
| If you're running a Unix-like operating system (not Windows) and get bind errors //only// when recycling servers: | If you're running a Unix-like operating system (not Windows) and get bind errors //only// when recycling servers: |
| sbbs: term 0001 Will retry in 15 seconds (1 of 2) | sbbs: term 0001 Will retry in 15 seconds (1 of 2) |
| |
| ... this is most likely because a TCP session is stuck in a ''TCP TIMEWAIT'' state (you can use ''[[man>netstat]]'' to verify this). The session will eventually time-out and close properly on its own, allowing the port to be re-bound at that time. To work-around this problem, you can either increase the ''BindRetryCount'' and/or ''BindRetryDelay'' values in your ''[[dir:ctrl]]/[[config:sbbs.ini]]'' file, or you can add the following line to the root section of your ''[[dir:ctrl]]/[[config:sockopts.ini]]'' file: | ... this is most likely because a TCP session is stuck in a TCP "TIME WAIT" or "CLOSE WAIT" state (you can use ''[[man>netstat]]'' to verify this). The session will eventually time-out and close properly on its own, allowing the port to be re-bound at that time. To work-around this problem, you can either increase the ''BindRetryCount'' and/or ''BindRetryDelay'' values in your ''[[dir:ctrl]]/[[config:sbbs.ini]]'' file, or you can add the following line to the root section of your ''[[dir:ctrl]]/[[config:sockopts.ini]]'' file: |
| REUSEADDR=1 | REUSEADDR=1 |
| |
