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
I guess from reading the code that Excon currently creates a whole new connection for each command executed on Docker?
In my app we send a lot of commands to many different Docker containers. I think it would greatly increase the speed if it continued to use the already open connection.
I have noticed the comment about bad requests and to solve that issue maybe we could nullify to connection instance variable when we register a bad request from the HTTP status code?
Or maybe it could be an option to cache the value? Then people who experience issues with bad requests could not opt in (or opt out)?
The text was updated successfully, but these errors were encountered:
I would suggest caching the value of the following method:
https://github.com/swipely/docker-api/blob/fc7ad57f77d59036dba9598ed52772778dc198c6/lib/docker/connection.rb#L29-L33
I guess from reading the code that Excon currently creates a whole new connection for each command executed on Docker?
In my app we send a lot of commands to many different Docker containers. I think it would greatly increase the speed if it continued to use the already open connection.
I have noticed the comment about bad requests and to solve that issue maybe we could nullify to connection instance variable when we register a bad request from the HTTP status code?
Or maybe it could be an option to cache the value? Then people who experience issues with bad requests could not opt in (or opt out)?
The text was updated successfully, but these errors were encountered: