Publish fails or is slow when publishing through a proxy, and connections are some being made directly from source to target bypassing the proxy (Doc ID 1497003.1)

Last updated on FEBRUARY 28, 2022

Applies to:

Symptoms

Publish from source to target fails or takes a very long time to complete. When failing, the following error is seen on source logs:

[2012-08-14 14:13:03,398] [fatwire.wem.sso.DynamicInjector] Accessing CASinfo using https://10.33.146.13:48461/cs/Satellite?pagename=fatwire/wem/sso/casInfo
[2012-08-14 14:16:12,406] [fatwire.wem.sso.DynamicInjector] Failed to access CASinfo: java.net.ConnectException: Connection timed out
[2012-08-14 14:16:12,417] [apache.commons.httpclient.HttpMethodDirector] Closing the connection.
[2012-08-14 14:16:12,418] [apache.commons.httpclient.HttpMethodDirector] Method retry handler returned false. Automatic recovery will not be attempted
[2012-08-14 14:16:12,418] [com.fatwire.logging.cs] MirrorQueue: Login failed, code=-3500, , Network connection disconnected[OpenMarket/Xcelerate/Admin/Publish/TargetSitePost#OpenMarket/Xcelerate/Admin/Publish/TargetSitePost]

We are publishing through a proxy, configured in publish destination argument PROXYSERVER & PROXYPORT. Same behaviour exists when proxy is configured in futuretense.ini properties cs.mirrorproxyserver & cs.mirrorproxyserverport. As this is a production environment, the source and target machines have been firewalled for security such that source cannot talk to target directly, and must go through the proxy. Source is allowed to talk to the proxy, and the proxy is allowed to talk to the target.

Also "netstat -an" and other monitoring has shown that there are attempted connections being made direct from CS source to CS target, that are not going through the proxy.

Changes

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.