• Bug
  • Status: Closed
  • 2 Major
  • Resolution: Fixed
  • gkeim
  • Reporter: tgautier
  • September 04, 2008
  • 0
  • Watchers: 0
  • July 27, 2012
  • September 08, 2008

Description

We have noticed on a slow connection, and with a lot of data, the lock profiler times out.

Here are the details:

1) Connection to server is relatively slow as it is on a remote WAN link 2) There are a lot of locks in the application 3) Setting stack trace depth to 1, usually results in a somewhat slow, but responsive lock profiler (e.g. hit refresh, get results) 4) Setting stack trace depth to 8, cannot get results from lock profiler. Once this happens, the lock profiler UI ceases to function in entirety 4a) Setting stack trace depth to 8, cannot get results from lock profiler unless the refresh button is pressed within a very short time.

Comments

Gary Keim 2008-09-04

There is an arbitrary timeout here. I think the correct solution is to remove the timeout and allow for cancellation. I’ll take a look at what that takes.

Gary Keim 2008-09-08

This change is now in 2.6, 2.7, and trunk.

Fiona OShea 2008-09-09

In 2.6.3-patch-3