Hi all, I thought I would followup on this issue.
When our Peoplesoft portal in hit multiple ST 302s are generated, usually 4 but occasionally 5, that is when the validation error occurs. We think this is the issue (from a co-worker): "This would not be much of an issue with an idempotent get request, but would cause the issue with CAS validation 500 errors we are seeing." Chromium Issue 317918: sending duplicate requests before and after 302 redirect “ The pre-render engine in Chrome is saying "You're visiting this page enough that I think it likely you're going to visit it very soon, let me pre-load it for you." Normally, I'd expect that to show up in cache the first time you load it and not hit the network after that. But the response is marked no-cache, no-store, so that's prevented from happening. Adding in the pre-render folks just in case we want to take either no-cache,no-store or a redirect as a signal not to pre-render. But I think that's a feature optimization, if we want to implement it, not a bug. “ We can’t prove this a Chrome only issue, or if it is limited to Peoplesoft. We can’t prove anything…. Anyway we have modded the CAS Java client to re-send a 302 ST validation request when we detect this. Using some fancy pants request params we count the retries to prevent an endless 302 loop… Additionally we have finally decided to have our Citrix Netscaler LB send a friendly 500 error page instead of sending a Java stack dump to the user. As always, any feedback / opinions / thoughts are appreciated. Can’t wait for first day of classes and not showing up on the radar. -Bryan -- You are currently subscribed to cas-dev@lists.jasig.org as: arch...@mail-archive.com To unsubscribe, change settings or access archives, see http://www.ja-sig.org/wiki/display/JSG/cas-dev
