Ask Your Question

Hight latency with down nodes

asked 2014-10-28 17:40:33 -0500

pskerrett gravatar image


I"ve been struggling with high latency on read & write requests when there is a down object node in my environment. OF course the latency is coming from the proxy server timing out when attempting to contact the down object node.

In my proxy.conf, I have conn_timeout = 0.5 & when the proxy server attempts to reach a node that is completely off-line (as in the IP address is not reachable at all), then I get a 0.5 sec delay from my clients.

The problem is that if I have 4 object nodes, and one of them is down, I get this delay on every 4th request. I am hoping there is a way to get the proxy servers to ignore this node if it fails X # of times & not attempt to contact it again for another XX seconds. To that end, I have been experimenting with the settings:

error_suppression_interval and error_suppression_limit

With various values, but it does not seem to have any affect on getting the proxy servers to stop attempting to contact an unresponsive node.

Are those the correct config items I should be working with, or am I missing something else somewhere.

Thanks in advance.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2014-10-29 07:06:12 -0500

pskerrett gravatar image

updated 2014-10-29 07:06:55 -0500

I finally found the answer buried in a github article.


Setting node_timeout = 10 and recoverable_node_timeout = 10 did what I needed it to do.


edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2014-10-28 17:40:33 -0500

Seen: 65 times

Last updated: Oct 29 '14