Dynamic Bypass

The Dynamic Bypass feature lets you configure the appliance so that specific errors from Web sites are explicitly not cached and subsequent requests to the Web sites are simply passed through for a specific time period.

For help configuring the Dynamic Bypass feature, see Enable Dynamic Bypass.


Why Dynamic Bypass Is Needed

The appliance follows the HTTP 1.1 specification, which stipulates that no 400- or 500-series errors are cached (except HTTP 410: Data Moved Permanently). If these errors occur, the proxy server passes them through to the requesting browsers.

Some Web servers erroneously pass cache control public headers along with the errors they generate. These headers override the default error handling by the caching system and cause the error to be cached on the appliance.

For example, if a Web server receives an unauthorized request for data and responds with an HTTP 403: Forbidden error accompanied by a cache control public header, the error will be cached. The appliance will then respond to all subsequent requests, including those from authorized users, with the 403 error.


How the Proxy Server Handles Dynamic Bypass

If dynamic bypass is enabled for a given error and iChain Proxy Services receives the error from an origin Web site, the site's URL is dynamically added to the bypass list and retained in the list for the period of time specified. All subsequent requests to that URL during the dynamic bypass time period are passed through to the Web site and do not go through the cache.

To continue with the above example, if the appliance has dynamic bypass enabled and the 403 error checked, it would not cache the error but pass it directly back to the browser. It would also add the site's URL to its dynamic bypass list. On subsequent requests to the same URL, iChain Proxy Services would bypass cache and passthrough requests to the Web site for the period of time specified. This lets the Web server determine whether to accept or reject access requests on a case-by-case basis.