• New Feature
  • Status: Closed
  • 2 Major
  • Resolution: Won't Fix
  • drb
  • Reporter: teck
  • May 07, 2007
  • 0
  • Watchers: 0
  • February 17, 2009
  • January 28, 2009


It’s not actually clear what the effect of this is…

The format of the session identifier for a websphere app looks like it has three components:

1) The first 4 characters are some sort of cache ID that might have something to do with websphere’s dynacache stuff 2) The real session identifier 3) a server identifier (aka cloneID)

For the time being we are ignoring component #1 and using the “default” cacheId of “0000”. This item is the reminder to get to the bottom of it. The terracotta class that generates session IDs should have a way to specify the cacheID, but there won’t be any instrumentation to hook it up to a “real” value under we understand what it means and where it comes from


Nathaniel Harward 2007-06-13

Antonio – with your work on the IBM HTTP server, do you think this is still necessary?

Antonio Si 2007-07-03

Yes, I still think we need to understand what is the cacheID used for and how/when is it being changed.

Fiona OShea 2007-07-03

moved out as this requires research

Tim Eck 2009-01-28

websphere AS support removed from 2.7 and future releases