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
There is a desire for certain applications to have connections time out more quickly than would normally be available using TCP keep alive timeouts.
A good mechanism for this would be to have a per-pipe activity timer, that closes a pipe if no activity is seen on it for a while. It's not clear that there is a good default for this other than to be disabled by default, as this is intrinsically an application specific thing. (In particular, there could be applications which don't have any activity for a while.)
Longer term (not this issue) we might consider adding a transport layer keep alive that uses shorter values. In fact, in NNG we do have such a thing for the ZeroTier transport. Making such a change would require changes to the protocol however, so it's out of scope for now.
The text was updated successfully, but these errors were encountered:
There is a desire for certain applications to have connections time out more quickly than would normally be available using TCP keep alive timeouts.
A good mechanism for this would be to have a per-pipe activity timer, that closes a pipe if no activity is seen on it for a while. It's not clear that there is a good default for this other than to be disabled by default, as this is intrinsically an application specific thing. (In particular, there could be applications which don't have any activity for a while.)
Longer term (not this issue) we might consider adding a transport layer keep alive that uses shorter values. In fact, in NNG we do have such a thing for the ZeroTier transport. Making such a change would require changes to the protocol however, so it's out of scope for now.
The text was updated successfully, but these errors were encountered: