GroupWise Post Offices do not accept newly created users

  • 7009118
  • 04-Aug-2011
  • 27-Apr-2012

Environment

Novell GroupWise 8

Situation

Situation:
When the GroupWise administrator attempts to create new users on most any Post Office in the GroupWise system, the user never gets to the Post Office database (wphost.db) and because of that, they cannot login to GroupWise.
 
Symptoms:
The new GroupWise users attempt to login to their Webaccess mailbox, they get the error - "You mailbox is unavailable, please contact your system administrator".
 
After reviewing verbose mode GroupWise MTA logs and POA logs, it is noted that there are no errors indicating a problem with accepting the admininstrative changes made by trying to create new users.  The changes would simply not get written to the Wphost.db .
 
It was noted that when no GroupWise administrators were making administrative changes that new GroupWise administrative files would get populated and then processed in the GroupWise Domains\Wpcsin\2\ directory, indicating an admin message loop.  These messages were opened and it was confirmed they were administrative changes from the past coming into the \2\ directory over and over again.

Resolution

NOTE 1: Be aware that this procedure WILL require you to shut down MTA's and POA's during a rebuild process that is just a part of the suggestions to follow. You cannot avoid this, there is no other way to fix this problem. There is an Admin message loop that is likely corrupting the WPHOST.DB files and causing them to grow in size.

NOTE 2: It is important to understand that it is NOT just a matter of REBUILDING the GroupWise Domains ( Wpdomain.db) and Post Offices (Wphost.db) from ConsoleOne. That is NOT enough to fix the problem. We must stop the Admin message loop that is corrupting the Post office databases. During this TOP Down REBUILD process I will describe below you MUST also rename 3 GroupWise Domain subdirectories in the process ( NOT the domain directory itself), they are /WPCSIN/, /WPCSOUT/, and /MSLOCAL/.
 
NOTE: 3: Continued from NOTE 4: You must also rename during this process 2 Post Office subdirectories, they are /WPCSIN/ and /WPCSOUT/. NOT the POST Office directory itself.
 
NOTE: 4: You must make a note of the location on disk of the GroupWise Domains and Post Offices, so you can efficiently change to the different directories and efficiently rename the listed sub-directories and shut down and restart agents efficiently so not to pro-long the time agents are down, You must have a drive mapping from the perspective of the Linux ConsoleOne workstation(s) you will be using, do all Domains and Post Offices.
 
NOTE 5: You should check that there are NO Domain database file (wpdomain.db) in any the Post Office directories and NO Wphost.db in the Domain directories, this can cause odd admin behavior. Check make a quick check and your done with this item.
 
NOTE 6: Make sure NO ONE is making admin changes in ConsoleOne, while you do this procedure.
*** So let's get started : Instructions start HERE ***

 Step 1 : Shut down ALL GroupWise components at the linux terminal prompt as the "root" user - "rcgrpwise stop". You will need to do this on ALL nodes of the cluster. So NO GroupWise agent, or gateway will be running temporarily, until instructed to do so.
 
Step 2 : As the "root" user, at a terminal , rename ALL /GroupWise Domain/wpcsin/, /wpcsout/, /mslocal/ subdirectories to another name like with a .OLD extension, in ALL GroupWise Domain directories
 
Step 3 : As the "root" user, at a terminal, rename ALL /GroupWise Post Office/wpcsin/, /wpcsout/, subdirectories to another name like with a .OLD extension, in ALL GroupWise Post Office directories
 
Step 4 : In ConsoleOne , be connected to the Primary Domain in the GroupWise view, and do TOOLS, GroupWise Utilities, System Maintenance , Rebuild Database.
 
Step 5 : Do the same procedure in STEP # 4 , but against ALL remaining GroupWise Domains.
 
Step 6: Rebuild each GroupWise Post Office database, being connected FIRST to the owning GroupWise Domain in the GroupWise view
 
Step 7: You are now ready to bring up the GroupWise system. AT a terminal prompt, as "root" do "rcgrpwise start" on all nodes.
 
Step 8: Test out to see if you now can duplicate the problem, by attempting to create a NEW uniquely named user in eDirectory and GroupWise as you have done before. Can you log into a GroupWise client after about 1 minute of sync time.
Using this exact procedure fixed this problem for a GroupWise customer.  This is a proven solution.