Mainframe - Crucial Role in Modern Enterprise Computing

1. 00:00Hello it's me again Wolfram Greis from the European Mainframe Academy,
2. 00:06and in the next video clips
3. 00:08we want to present a real impressive success story from Walmart
4. 00:13in the United States.
5. 00:15Walmart is one of the biggest IBM
6. 00:18mainframe customers worldwide and the walmart adopted the
7. 00:23mainframe in 1975 and it has always played a very
8. 00:30important part in the IT environment of Walmart and of course developed during
9. 00:37this more than forty years.
10. 00:41And there's another
11. 00:44youtube video clip from Randy Frerking, distinguished engineer at Walmart
12. 00:50and he is interviewed by IBM and he tells in this interview
13. 00:55that Walmart processed in 2015, 500
14. 00:59million transactions a day and 150 million of these transactions
15. 01:04are in between web services. So very new technology to access
16. 01:12the data and transactions on the IBM mainframe.
17. 01:18There's a very interesting website I don't know if you're or you are
18. 01:22familiar with it. It is the IBM redbook website. Just point your browser to redbook.ibm.com
19. 01:28and you get a lot of information about
20. 01:32IBM platforms. Not just the mainframe but of course also about the IBM mainframe.
21. 01:40And then why do I show this because
22. 01:45Walmart contributed to some very new redbooks, starting in
23. 01:502016 with redbook with the title Creating
24. 01:55IBM z/OS Cloud Services, followed in April 2016
25. 02:01with redbook how Walmart became a cloud service provider with IBM
26. 02:07CICS. And a third one quite new
27. 02:11in February 2018, CICS and the
28. 02:16IBM CICS and the Coupling Facility Beyond the Basics. Because we talked
29. 02:22during this MOOC about CICSplex,Parallel Sysplex and so on. And
30. 02:28this Walmart infrastructure is really a very impressive implementation
31. 02:34of such a clustering technique.
32. 02:40These out of a series from twelve video clips complementing the redbooks
33. 02:46mentioned. We show here the first with the introduction and
34. 02:51the last with the summary. All the others if you are interested in,
35. 02:55you can get from this redbook website. Let's start with the
36. 03:01introduction.
37. 03:10I'm Randy Frerking, distinguished engineer at Walmart
38. 03:13and I'm Rich Jackson, principal systems engineer at Walmart.
39. 03:17In 2007, Walmart averaged two hundred million
40. 03:20transactions a day in CICS primarily in thirty to seventy
41. 03:24MQ and btm sna access
42. 03:27In 2009, we introduced web services primarily
43. 03:30SOAP and primarily used as a data access layer, and by two thousand
44. 03:33and eleven we had grown to five hundred million transactions a day.
45. 03:37In 2013, we introduced rest services and developed
46. 03:40our first cloud services- a key-value database, a caching service
47. 03:44and a unique ID generator. We also created a dedicated
48. 03:48CICSplex called our CloudPlex.
49. 03:51You don't like that term, do you? I do not like that term.
50. 03:53You get used to it. And now we average about seven hundred million transactions a day.
51. 03:58We were able to achieve this volume by leveraging
52. 04:01coupling facility and fully utilizing Parallel Sysplex.
53. 04:05Some of the traditional components of Parallel Sysplex include things like grs,
54. 04:09group buffer pools are just two check points.
55. 04:12Most recently CICS has added new features to the coupling facility,
56. 04:15such as the Named Counter, Coupling Facility Data Table, Shared Temporary Storage Servers
57. 04:21and record level sharing. In this video course we'll share some
58. 04:24things that we've done with all of these CF technologies. Thank
59. 04:27you for watching.
60. 04:38Now let's show the summary.
61. 04:51In this video series, we've covered a number of specific CF technologies that
62. 04:54can be used in interesting ways.
63. 04:57These features can be combined to build reliable, available and highly scalable
64. 05:01application solutions.
65. 05:03We'll briefly cover a couple of examples where we've taken
66. 05:06some of the specific CF technologies and combine them into something greater.
67. 05:11The first example would be Sequencing captured updates in real
68. 05:15time in a fluid manner.
69. 05:18In our particular problem,
70. 05:21we have an application that needs to synchronize updates from that application
71. 05:26to another application.
72. 05:28The problem with that is, after the data has been captured it needs to be
73. 05:33transformed. And during the transformation process, which does take time
74. 05:37the data gets out of sequence.
75. 05:39So the challenge is reapplying the updates to the target system,
76. 05:43re-sequenced into the order that they actually occurred.
77. 05:46We use several CF technologies to create the solution. One of
78. 05:50them was VSAM RLS, which we used to house shared control information.
79. 05:56We also use CFDT. We stored a number of things in there, such
80. 05:59as additional control information, sequence numbers for the update
81. 06:03and we also used it for caching. We also use GRS for coordination
82. 06:07of information between parent and child task.
83. 06:10As well as establishing control in our transaction environment
84. 06:12by using it as global transit class.
85. 06:15In our next example, we had a requirement to process a one to
86. 06:19many parallel processing service.
87. 06:22For this particular problem, the scenario involved a single incoming request
88. 06:28that triggered numerous parallel outbound requests to remote sources.
89. 06:34The information from those remote sources needed to be combined,
90. 06:36aggregated and returned to the requester. To create the solution,
91. 06:40we use several CF components. One of them was using shared temporary
92. 06:44storage for task coordination between the primary and multiple secondary tasks.
93. 06:48Yes and we also used GRS to serialize activity against those TSQs.
94. 06:53We also use CFDT for caching information that was retrieved
95. 06:56from the secondary task provided to the primary task.
96. 06:59And that was very important in this case because the remote hosts
97. 07:02involved in this situation were network attached devices, so
98. 07:05there were significant latency.
99. 07:08More examples like these as well as additional details on everything
100. 07:11we've talked about in this video series are available in the redbook.
101. 07:14Thank you for watching. This was the introduction
102. 07:25to the Walmart success story. For questions and remarks
103. 07:30concerning our video clip, please use the discussion forum
104. 07:34on the OpenHPI platform. Thank you very much for attending and
105. 07:42we will see us in one of the next video clips.

To enable the transcript, please select a language in the video player settings menu.