IBM Confidential: System/360 File Organization

My name is Pearson Wood. I'm with the data storage planning group data storage planning in Poughkeepsie, New York. This presentation is one of a series which is intended to present to you the systems engineer and salesman some of the advantages of systems 360. This particular presentation has to do with file organization for the IBM systems 360. The first thing that we must take into consideration is what is a file. We can take a very simple sentence, "File the file in the file." And it uses the word file three times, but it has three different meanings. The first word file meaning some action, do something. The second word file having to do with the content. And the third word file having to do with a device. With this particular presentation, we are concerned with the second word file, that is the content. So, again, file organization for systems 360. Actually, we can't discuss file organization without talking about data processing as well. And we have some relationships which should be displayed which can explain more about organization and processing. We have four cases shown here. We have sequential data. We have random data. Against this data, we are processing either sequentially or randomly. And any application which you will be doing with the System/360 will fall into this category, either sequential data, random data, sequential processing, random processing. One of these four particular cases. Processing and data organization. We have shown here some data files. And in this particular case, all of these data files are sequentially organized data, starting with the number one going to the highest end. And in this particular first case, we're going to talk about sequential processing against a sequentially organized data file. And that means that we are coming in with our transaction data, going to record one, record two, record four, six, seven, and so on through the file. We can compare this to the way that we would might process a payroll. That is, we would pass the data file for every employee, and we would prepare a check for every employee. Our second example is again a sequentially organized data file, and we now have skip sequential processing. And what do we mean by skip sequential processing? Well, basically, it's this. We're still processing in a sequential manner, but we are skipping through the file. This is the way batch processing is done. First of all, record four, record six, nine, on to 13, skipping through our data file. This is very commonly how batch processing is done. We can have large batches of data or small batches of data. If we were to use magnetic tape for small batches of data, however, we must take into consideration that we would have to pass the entire data file in order to get just these few transactions processed. Whereas with a direct access storage device, we would just take into consideration the number of hits that we have, multiply them by the time that it takes to process a transaction, and we will know how long it will take us to process these transactions. In this case, again, another sequentially organized data file, and we are randomly processing against this file. Record six, back to record two, to 10, 14, back to seven. This gives us a very definite advantage. We can now use IBM teleprocessing equipment as well. We can teleprocess our data in, and we can handle transactions as they are received. We have the ability right on the spot to make inquiries into a data file. These last two examples have to do with randomly organized data files. I have, again, shown two data files, same length as before, but the data is random within it. But, I can process sequentially against this data file. I can come in with record one, go to record two, three, four, five, and so forth. So, I have the ability to do sequential processing against a randomly organized data file, as well. This last example, a mixture. Random processing, randomly organized data. So, IBM does have the data storage equipment to handle all of these categories of processing and data organization. We are used to sequential processing and we're used to data processing in a sequential manner from our card and magnetic tape days. This is the way we processed with our unit record equipment and this is the way we processed with tape equipment. When we came out with the 305 RAMAC system, we had the ability to have random data organization and random processing. With sequential data organization and sequential processing, we have the ability to get out reports. Reports are primarily done in a sequential manner and people like to receive them in that manner. Batches of information are handled sequentially. When we process randomly against a sequential file, we have the ability to handle inquiries and we can take care of priority items. That is, we can handle transactions as they are received, we can act right now. Most data file organization is sequential. It just happens to way be the way that we live. Our education comes out that way. We are taught to count on our fingers from 1 to 10. We do our ABCs starting with A and going all the way through Z. We operate in a sequential manner. Well, does this mean that we have to continue on this way? Well, maybe yes because we have to take into consideration that mail delivery or whatever means that we have for receiving data is probably sequential and probably will continue for some time. So, I've said right now that most data file organization is sequential. This is true. It will probably be true for some time to come. Well, with the sequentially organized tape file, what kind of processing can I do? Well, I can have sequentially processed data, as I pointed out before. We can have batch processing. Although, I will probably want to have large batches of data, rather than small batches of data. For instance, in a payroll situation, if I have 11 changes which have to be made, let's say 10 rate changes and one address change. In order to do them in a batch basis with magnetic tape, I would probably not want to handle those 11 transactions at one time. I'd probably wait until I had a lot of them, because I would have to pass my entire file in order to do it just those 11 transactions. With a direct access storage device, I could do it right now. I probably won't be doing direct processing, and I don't have inquiry ability with magnetic tape. With a sequentially organized direct access storage device, I can have sequentially processed data. I can have skip sequentially processed data. I can have batch processing, and I really don't care if I have large batches or small batches. I can have random processing, taking advantage of our teleprocessing equipment, and I have inquiry ability. Teleprocessing, in information or inquiry right on the spot uh where my computer is. An index is a very important tool to use with a data storage device, and it has several advantages. Primarily, what an index does for us is permit a common organization for all of our direct access storage devices. That is that I can prepare a program and let's say that I want to use a 2311, a 2302, or a 2321 data cell, I can prepare one program which is good for installation, programming, or conversion with all of these devices. This means that I have very convenient growth as well because now with one program with no change or very little change to a program, I can capitalize on an improvement that IBM might make in a product. So, I can go from one device to another very easily. Device changes again. If I want to go from the 2321 where I have up to 400 million characters of information online and I decide that well, I had just have to react a little quicker to uh transactions. I've got more processing coming in and I may want to switch to a 2311 type device, I can do it very easily. The same program. So, I have very convenient growth. Indexing allows all types of data processing activity as well. I can have sequential processing, skip sequential processing, random processing, random inquiry, teleprocessing, all of these advantages all on the same data within one data file through the use of an index. Well, just a very simple explanation what is an index? Well, an index is an ordered reference list of the contents of a file and it's very similar to the index of a book. A book is a sequentially organized data file. We have an index which takes us to the page. It tells us where to go for our information. Another example might be a road map. And this gets us into a double level of indexing. For instance, if I were looking for the city of Flint in Michigan, I would go to an index and the index would tell me that the state of Michigan is found at K4. And I go to K4 and there's Michigan. This will take me to another index and the other index tells me that going to Flint, Flint will be found at D5. And I go to D5 and I found Flint, Michigan. Now, this is an example of a double level of index. There are very Well, there are several index types to be used, some of them quite sophisticated, but they all use the same basic principle. An index on a data file. In this particular case, I have a data file with information. It's divided up into tracks. And in the first track, I have shown an index. If I were looking for record number five, I would go to my index and I would find that number five is larger than one, it's smaller than 106. So, I must be on track two for record five. I go to track two and there I find record five. If I were looking for record number 109, I would go to my index. I would find that it's larger than 106, smaller than 211. I must be in track three. By going to track three, I go in and I can find record 109, the one that I want. The port of our devices, we have full programming support. We have full support programming support in the area of device independence. This means that a program that is written for one data storage device can be used on another data storage device. So, we have interchangeability through programming. We have COBOL and multipurpose programming language file statements. Systems residence, sort utility programs. In addition, we have full data management support. So, the programming support from IBM is again another tremendous advantage which no other competitor has to offer for their storage devices. We have some very interesting information which is available right out of your sales manual. And I think this is a little history right here that will point out some of the tremendous advantages and some of the experiences that we've had with our data storage devices and our in particular our disk storage. I want to point out now the fact that what has this experience done for us? What have we learned through this experience? Well, early we came out with the 305 and the 305 had a rate of cost per million characters that went something like this. This being cost, this being capacity. And the 305 came out about 19 uh 56. Then we came out with the 1405. And the 1405 had a range cost per uh character, let's say, in a cost per million character range that was went something like this. So, we brought down our costs. Then as time went on, we came out with another device. And we were able to again take advantage of our experience and bring down the cost of our equipment. Then we went into another device. And so again, we came down with our cost per million characters. Till we reached 1964, and we came out with the 2321 data cell. And here our cost per million characters starts here and just keeps going down. So that what we have done is taken advantage of our experience. We've brought down our cost per million characters, and this is just the cost of storage. At the same time, we have gone up with our performance rates. This is, I think, something that's very significant and should be brought out for the use of our salesmen and brought to our customers. We have some unique 360 advantages with our data storage devices that you should know about. We have something new, format as you load. Prior to 360, the formatting was done in advanced to predetermined requirements. The programmer had a lot of work to do. He had to lay out his records and have his formatting done. He had to do this before he could load a program. With the 360, the formatting is done automatically as you write new records. So you just lay in your information and the system will take care of formatting for you. In the area of keyed and non-keyed records, in order to use teleprocessing equipment, and if I were to bring in several different types of information with teleprocessing equipment, I would be coming in into my computer and then store this information a tape drive for every type of information that I were gathering. Or I could store it all on one tape, but before I could use it, I would have to sort the tape before I could use the information. Now, I can come into a computer, store the information on a disk, and I just basically segment my disk, and when I want to process the information, I can go directly to that area of the disk, get my information, and process it. This is very convenient for output reports as well as teleprocess the same. When the system is ready to print and a printer is free, I can go to an area on the disk, get out my information, and print my report. Another advantage of the System 360. Another unique advantage, symbolic assignment of storage. And primarily, what this does for you, it allows me the ability to store on one disk pack or on a data cell some information. I can write on one, I can take it to a shelf and store it, and when I bring the information back to be processed, I don't have to go back to this particular drive in order to process. I can take it over here if I want to. I can go to any of them. As a matter of fact, I can go to another system if I want to. We've got the products which cover a an entire range of data rates, access times. We can handle large records, small records, we can handle them all. We've got the programming support for our equipment. The rest is up to you. It's based on knowledge of your customer's records and the way that he wants to do his data processing. We've got the equipment to do the job. Good selling. That's it. A member of the audience has been given the responsibility for answering any questions you may have on this presentation.