tag:blogger.com,1999:blog-1812985003737084262.post737023376652876417..comments2009-03-24T16:45:45.909-04:00bxiahttp://www.blogger.com/profile/13262666423559414651noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-1812985003737084262.post-69150686395674850472009-03-24T16:45:00.000-04:002009-03-24T16:45:00.000-04:00When WinDBG loads the dump file, it displays the system and process up times, for instance,<BR/>System Uptime: 10 days 3:17:01.859<BR/>Process Uptime: 10 days 2:46:49.000.<BR/><BR/>The !runaway extension command displays thread elapsed times, such as<BR/><BR/>0:013&gt; !runaway 7<BR/>...<BR/>Elapsed Time<BR/>Thread Time<BR/>0:9f8 10 days 2:46:48.317<BR/>2:9d8 10 days 2:46:48.270<BR/>1:670 10 days 2:46:48.270<BR/>3:858 10 days 2:46:47.708<BR/>4:7ec 10 days 2:46:47.692<BR/>5:994 10 days 2:46:46.833<BR/>6:878 10 days 2:46:46.817<BR/>7:478 10 days 2:46:44.301<BR/>8:ebc 10 days 2:46:44.161<BR/>9:ed0 10 days 2:46:44.098<BR/>10:8b0 10 days 2:46:43.958<BR/>11:8a8 10 days 2:46:43.911<BR/>12:804 10 days 2:36:43.301<BR/>13:834 10 days 0:20:23.708<BR/>14:f34 10 days 0:11:55.114<BR/><BR/>The question is when the up/elapsed time starts -- when a process/thread is created or when a debugger attaches to a process.<BR/><BR/>I attached the debugger to the server and investigated the issue within ONE day, therefore, the up/elapsed time is from the beginning of the life cycle of a process/thread.bxiahttp://www.blogger.com/profile/13262666423559414651noreply@blogger.comtag:blogger.com,1999:blog-1812985003737084262.post-36154753727208594062009-03-18T00:40:00.000-04:002009-03-18T00:40:00.000-04:00Prefer a named event over an unnamed event. Since it is more debugger-friendly.bxiahttp://www.blogger.com/profile/13262666423559414651noreply@blogger.com