[GitHub] trafficserver pull request: Ts 3656

2015-06-16 Thread asfgit
Github user asfgit closed the pull request at: https://github.com/apache/trafficserver/pull/215 --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature

[GitHub] trafficserver pull request: Ts 3656

2015-06-15 Thread SolidWallOfCode
Github user SolidWallOfCode commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-112159037 My understanding is it is a plugin that enables redirection following for the request but the request is from a user agent. I am aware this is not complia

[GitHub] trafficserver pull request: Ts 3656

2015-06-15 Thread sudheerv
Github user sudheerv commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-112150121 This is not a real UA we are talking about. It's a plugin IIUC. If you are talking about real UA, fyi, what ATS does with internal redirect is non-stand

[GitHub] trafficserver pull request: Ts 3656

2015-06-15 Thread SolidWallOfCode
Github user SolidWallOfCode commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-112147592 It's not really feasible to ask the user agent to change behavior. In my view the bottom line is that ATS does this, it's an incompatible change to not do

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread sudheerv
Github user sudheerv commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-110417011 Can you ask the customer/user not to use redirect with POST (with the supporting material from above) I believe I've done that on more than one occasion in the p

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread shinrich
Github user shinrich commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-110416286 The existing code does resubmit the body if the enable_redirect is set statically or set early enough in the life cycle of the transaction. This change adjust t

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread sudheerv
Github user sudheerv commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-110415828 AFAIK, we have always maintained (within our prod at least) that the behavior of redirect follow w/ POST is not guaranteed. As such, the current code of not resu

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread shinrich
Github user shinrich commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-110415110 The existing code goes to some effort to support post redirect. That logic has been there since before I got involved. It is a broader discussion if we want to

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread sudheerv
Github user sudheerv commented on the pull request: https://github.com/apache/trafficserver/pull/215#issuecomment-110394004 Do we claim to support redirection for POST? POST is not an idempotent method and AFAIK, it is highly unreliable and not recommended to use redirection with POST

[GitHub] trafficserver pull request: Ts 3656

2015-06-09 Thread shinrich
GitHub user shinrich opened a pull request: https://github.com/apache/trafficserver/pull/215 Ts 3656 Changes to allow activating follow redirection in send server response hook. Originally this would not work for the post case, because the buffers storing the post data for redirec