Hi List,

 I am maintaining a few continuous replication connections, that are
tunneled through SSH tunnels.

 However, it seems that when an SSH tunnel process dies, or the SSH
server is unreachable or stops responding, when I try and start the
same replication connection (I know couchdb re-uses them and
identified them according to the parameters) it returns ok, instead of
reporting error (that should be propagated to couchdbkit which I'm
using for my works).

 Is there a way to set the timeout or other settings such that
connections that are not reaching the other end will drop quickly?
Could it be that CouchDB will not know a continuous replication
connection is lingering until new changes are done on the db and need
replicated to the other end of the continuous replication connection?

A lead on this would be highly appreciated!

Many thanks,

Sivan

Reply via email to