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
when using the Sender in cluster mode it seems like the python client always try to check all nodes from the beginning till a match is found every single time. this is a waste.
is it possible to store the latest active_node per cluster and try onto that one and check all others only in case of timeouts ?
The text was updated successfully, but these errors were encountered:
So, the main point here is that the active_node is stored in the object Sender. Using a single Sender object allows you to keep the latest active_node and make the library connect to this node first.
You probably used several Sender instances that resulted in ordering cluster nodes every time from scratch.
I hope my explanation will help you deal with the problem mentioned. Whatever, feel free to create another issue if the problem still exists.
when using the Sender in cluster mode it seems like the python client always try to check all nodes from the beginning till a match is found every single time. this is a waste.
is it possible to store the latest active_node per cluster and try onto that one and check all others only in case of timeouts ?
The text was updated successfully, but these errors were encountered: