You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
IIUC, the current logic for TCP's timeout is whenever some traffic can be read from tun, the timeout will be reset. However, when doing long-time write to TCP(i.e. write to tun), we do not reset the timeout value, which will cause the timeout when poll_read is invoked. I think the timeout logic should be similar to what has been done for UDP. Whenever there is some traffic on the stream, the deadline should be extended.
The text was updated successfully, but these errors were encountered:
IIUC, the current logic for TCP's timeout is whenever some traffic can be read from tun, the timeout will be reset. However, when doing long-time write to TCP(i.e. write to tun), we do not reset the timeout value, which will cause the timeout when
poll_read
is invoked. I think the timeout logic should be similar to what has been done for UDP. Whenever there is some traffic on the stream, the deadline should be extended.The text was updated successfully, but these errors were encountered: