ESP Workload Automation

Expand all | Collapse all

Test Strategy for Implementing ESP

  • 1.  Test Strategy for Implementing ESP

    Posted Nov 13, 2009 06:11 AM
    We are implementing ESP at our shop, a large mainframe shop in the financial services area.   We are developing a test strategy and wondering whether there is someone out there willing to share their document and experience.   Thanks.


  • 2.  Re: Test Strategy for Implementing ESP

    Posted Nov 13, 2009 07:28 AM
    Mark, I would imagine that you are using professional services to help you with the conversion.They have many validation tools as part of that process that work well.When we converted, we also made sure we took advantage of the  advantages that ESP brought to the table.There were some things that when they converted, were real ugly, so we reworked them.  High level for ours was.Install and test in our  play ground, then DEV and then Prod.External communicators to your existing product are key.We tracked what the other scheduler was doing into ESP so it helped us validate better our hist file size etc.  Don't know if this helps, let me know any other information you might need.


  • 3.  Re: Test Strategy for Implementing ESP

    Posted Nov 13, 2009 07:48 AM
    Thanks for your quick response.   We are trying to document the test cases and steps, the pre and post conversion validation that we will do alongside the CA team.   We just want to learn form the experience of others.   Did you document your process?   Can you share the document?  


  • 4.  Re: Test Strategy for Implementing ESP

    Posted Nov 13, 2009 08:08 AM
    I don't have any consolidated document to share.Identify and look at all your external communicators closely. Document  and publish to  your group, if  1 then convert to x, if 2 then convert to  y etc.    Identify and make sure you understand your data set triggers. Types, data, connect direct, via external ...Any special  JCL controls depending on what tool you are using.  What tool are you converting from?


  • 5.  Re: Test Strategy for Implementing ESP

    Posted Nov 13, 2009 08:18 AM
    CA7.   I am part of a group on the Application side trying to protect our systems through this process.


  • 6.  Re: Test Strategy for Implementing ESP

    Posted Nov 13, 2009 11:04 AM
    Ok, from an applications perspective.  Test and validate your interfaces, u7svc, bti, cci etc.  Internal dataset triggers.  #JI/JO type JCL cards.    The tools will validate the schedules, so let them, the stuff above is where issues can show up.    In CA-7 Look Backs/ Lead Times can be a chore at times, in ESP can solve 95% of them but the conversion will be same, 8 hr look back in CA-7, 8 hr scope in ESP.  Look to convert to SCHEDULED('TODAY')   , Maybe even add a APPLID(MYAPPL) etc.    Look to take advantage of what ESP can do, conditional processing, generate variables, make schedules/jcl generic.    The more you look at it, the more you will get opened to many more possibilities available.


  • 7.  Re: Test Strategy for Implementing ESP

    Posted Feb 15, 2010 12:21 AM
    Hi,  We are looking to make the same move from CA7. How did your conversion go?Was it on time, Did you find any gotcha's, Did you go Big Bang etc. Is the product delivering the functionality you were looking for?        Any feedback will be appreciated      MA


  • 8.  Re: Test Strategy for Implementing ESP

    Posted Feb 15, 2010 02:41 AM
    We are halfway through our environments and we find new issues every time.   We continue to develop our pre and post conversion verification procedures.   Get your application people involved and responsible for the verification.   Have them check pre-conversion that everything that was scheduled in CA7 is scheduled the same in ESP - predecessors, successors, etc.   Post-conversion have them check that everything expected to run did so and the correct number of times and at the right time.What industry are you in?


  • 9.  Re: Test Strategy for Implementing ESP

    Posted Feb 15, 2010 03:20 AM
    We are in  retail  and are  being told that the conversion will take around 4 months as we dont have a large batch environment. From what I can derive from  your reply, the biggest issue appears to be the actual conversion process and its verification.  I take it you are manually going through the returned  converted code on a job by job basis.  What would you say is the success rate of the conversion process?   I accept that each site is different but I am looking for a rough figure to put into our project estimates.                    We are looking to do this over the next few   months and would be happy to have an discussion off post.        


  • 10.  Re: Test Strategy for Implementing ESP

    Posted Feb 16, 2010 02:33 AM
    We are a financial services company with no room for error on the conversion process.   I am on the applications side, not a member of the conversion team.   In this sense, I am a user, not a technician.   That is why I am concerned with verification by application owners.   We are converting from CA7 and involvement from application owners is building with each successive environment.   We have had to develop our own means of verification.Our conversion has taken much more than four months already and we have two more environments to do.  


  • 11.  Re: Test Strategy for Implementing ESP

    Posted Feb 22, 2010 09:49 AM
    Hi everyone. I'm Andre and this is my 1st post in this forum and I'd like to share my experience.    We had to migrate all our applications from Espresso 4.4 to CAWA 11.1 SP1. Once we have about 200 applications and more than 30 agents, we  performed the migration in different moments with specific groups of applications.  We are  an insurance corp. and there is no room for errors.  We used the CA  migration tool that converted  the artifacts  without any problem.  During the migration process we should be able to manage different applications on Espresso 4.4 and CAWA 11.1 SP1 at the same time. Finally, we finalized the migration process last week with no downtime.  Regards,Andre