(file) Return to readme.pulloperations CVS log (file) (dir) Up to [Pegasus] / pegasus

Diff for /pegasus/readme.pulloperations between version 1.1.2.2 and 1.1.2.6

version 1.1.2.2, 2013/06/03 22:34:01 version 1.1.2.6, 2013/06/12 00:56:21
Line 2 
Line 2 
  
 STATUS STATUS
  
 <<< This section is being maintained during the review and checkin process  <<< The TODO section is being maintained during the review and checkin process
 to keep track of problems, errors, notes, etc.  Must be deleted before to keep track of problems, errors, notes, etc.  Must be deleted before
 checkin to head of tree. Please feel free to add notes, etc in this checkin to head of tree. Please feel free to add notes, etc in this
 section as you review/test.>>>>>> section as you review/test.>>>>>>
  
   NOTES On working with task branch.
   
   Merge out Process
   
      To keep our TASK branch in sync with the current head of tree we need
      to do a regular merge out.  the TaskMakefile contains the makefile
      procedures to do this efficiently.  NOTE: Following these procedures is
      important in that you are merging out new material each time you do
      the merge out.  If you were just to repeatedly merge out, you would be
      merging previously merged changes a second time causing a real mess.
   
       Start with new directory and put TaskMakefile above pegasus (needed so you
       have this file for the initial operations.
   
         make -f TaskMakefile branch_merge_out BNAME=PEP317-pullop  ## takes a long time
   
      This checks out current head, merges it into task branch and sets tags
      for the mergeout.  Note that at the end of this step this work is
      part of the TASK... branch.
   
      NOW check for conflicts, errors, etc. that resulted from the merge.
      Look for conflict flags, compare the results (I use linux merge as a
      good graphic compare tool) and build and test. When you are satisfied
      that the merge out is clean, you can commit the results to the TASK...
      branch
   
      To commit the work to  this into Task branch
   
         make -f mak/TaskMakefile branch_merge_out_commit BNAME=PEP317-pullop
   
     or manually commit and finish as follows
   
       cvs commit
       make -f mak/TaskMakefile  branch_merge_out_finish BNAME=PEP317-pullop
   
   ## This last step is important since it cleans up temporary tags to prepare
      you for the next checkout
   
   COMPARE TASKBRANCH WITH HEAD
   
   In a new pegasus work space do same as above for merge out.
   
   make -f TaskMakefile BNAME=PEP317-pullop
   
   This produces a result which is all of the head merged into the branch.
   A diff of this is all the new changes to the head of tree that you will
   include into the merge.
   
   
   TODO list:
      1. Binary operation from OOP.  Need to add counter to binary
         protocol to be able to count objects in response. Generates
         warnings in things like messageserializer and does not work with
         OOP right now.
      2. OpenExecQuery - Code is incorrect in that it used InstancesWithPath
         where the spec is instances with no path.  Need new function to wrap
         getInstanceElement(withoutPathElement) in XmlReader.  Note that
         Alternate is to put flag on InstancesWith Path to say no path
      3. Code for Pull part of OpenQueryInstancesRequest a) should be part of
         the common CIMOperationRequestDispatcher execCommon code.
      4. The changes to WQLCIMOperationRequestDispatcher and CQL... for handling
         pull not completed so we feed the responses back to the EnmerationContext
         queues
      3. Lots of minor TODOs, etc.
      4. External runtime variables. Decide this as part of PEP. The variables
         exist in CIMOperationRequestDispatcher but not in CIMConfig.  The primary
         ones to consider are:
         a. System maxObjectCount.  Setting some maximum size on what a pull
             client can request (i.e. the maximum size of the maxObjectCount on
             Open... and pull operations.
         b. Pull interoperationTimeout (max times between operations). This is
             the maximum number of seconds on the operationTimeout parameter of the
             Open operations
         c. Maximum size of the responseCache before it starts backing up
             responses to the providers.
      5. Decision on EnumerationContext timeout (separate thread or just
         checks during other operations). Can we, in fact really keep the
         enumeration context table and queue under control without monitoring
         with a separate thread. We must monitor for:
         a. Client operation that stop requesting (i.e. inter operation time
             exceeds operationTimeout). Note that if it simply exceeds the time
             the next operation does the cleanup.  The issue is those clients that
             simply stop and do not either close or go to completion.
         b. We should protect against providers that no not every finish delivering
             or take to long between deliveries.  This does not exist in Pegasus
             today
      6. Clean up code in Dispatcher. The associators code is still real mess
         and the pull code is in a template.  The Pull code is good now but
         must be duplicated.  Look at creating new CIMMessage CIMPullResponseMessage
         so that we can have common code.  Everything is the same except what
         goes into the CIMResponseData so it is logical to have completely
         common processing
      7. Extension to avoid double move of objects in CIMResponseData (one
         into enumerationContext queue and second to new cimResponseData for
         response.  Want to avoid second move by extending Open/Pull response
         messages to include count and CIMResponse data to count objects out
         of queue when converting (avoids the second move).  Big issue here
         with binary data since need to extend format to count it.
      8. Still using templates, etc. in  code in the Dispatcher. This is for
         all of the open operations where there is a lot of duplicate code
         and the pull operations that are 99% duplicate code (in a single template)
   
   Fixed for 9 June CVS update
      1. Cleaned up code for OpenQueryInstances.  Note that this is incomplete.
         No support in WQL or CQL Operations
      2.
   
   What was fixed for 5 June checkin.
      1. Extended ResponseTest MOF for for both CMPI and C++ subclasses
      2. Fixed issues with pullop.
      3. Fixed temp issue with CIMResponseData size by putting in mutex. That
         is not a permanent fix but it gets around issue probably in the control
         of the move logic that meant counts were off.
      4. Fixed issues in Dispatcher so that associator code works. Still messy
         code in the dispatcher.
      5. Changed name of Enumerationtable.h & cpp to EnumerationContextTable.*
      6  Changed name of ResponseStressTest module, classes, etc.
   
   TAG: TASK_PEP317_5JUNE_2013_2
   
 2 June 2013 2 June 2013
  
 Issues  - KS Issues  - KS

Legend:
Removed from v.1.1.2.2  
changed lines
  Added in v.1.1.2.6
No CVS admin address has been configured
 Powered by
ViewCVS 0.9.2