https://bz.apache.org/bugzilla/show_bug.cgi?id=69405
--- Comment #3 from Graham Leggett <minf...@apache.org> --- (In reply to Christopher Schultz from comment #2) > The information is discoverable. If the information is discoverable, the error message needs to include information on how to discover it. > Since this error would really only be expected in development, having a > method that re-scans the filter chain looking for async support would be an > option rather than building that information for every request and then > discarding it every time. This code has long since been developed, and has not been changed recently, the idea that "it shouldn't happen" isn't true, as it just happened. The most likely source of the problem is a vendor update to the Tomcat RPMs that have enabled something internally. What should be happening is that on the error path (and not the critical path for the reason you've already mentioned) we walk the request again and determine which component has returned false so that component is printed in the error message. We need to be nice to ops. -- You are receiving this mail because: You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
