c revised 05/19/86 rem. changed 64bc to 21598 and GDP to EDP C*********************************************************************** C COPYRIGHT ATLANTIC RICHFIELD COMPANY 1991 * C*********************************************************************** c revised 09/30/86 cmp. removed references to SEL c revised 12/09/87 rem. document "noextend" parameter. c revised 10/06/88 cmp. removed references to mass storage c revised 04/12/89 esn. added in 'summary' for jgparm. CUPAGE INTRO CUINDX INTRODUCTION TO THE SPARC SYSTEM 00000100 CU Introduction 00000200 CU 00000300 CU The SPARC (Seismic Processing Atlantic Richfield Company) system 00000400 CU is the primary software system used by Exploration Data Processing.00000500 CU SPARC serves the needs for production processing and analysis as 00000500 CU well as research and development. 00000600 CU 00000800 CU Several user oriented features make SPARC easy to use. These 00000900 CU include simplified user management of seismic data, uniform 00001000 CU data card formats, plus allowing the user to form processing 00001100 CU jobs without using computer job control language. 00001200 CU 00001300 CU The SPARC system has a flexible structure which is continuously 00001400 CU being adapted for new seismic data processing techniques. New 00001500 CU programs and new features keep SPARC a state-of the-art system. 00001600 CUEND 00001800 CU Main Features 00001900 CU 00002000 CU SPARC provides the Exploration Data Processing professional with 00002100 CU a large variety of processing tools. These include basic 00002200 CU facilities for filtering, deconvolution, normal moveout correction,00002300 CU stacking, etc. Also included are advanced facilities for automatic00002400 CU surface consistent statics, 2-D/3-D finite difference migration, 00002500 CU depth migration, over four special deconvolution methods, etc. 00002600 CU 00002700 CU SPARC jobs can be composed of any sequence of processes which 00002800 CU follow sound geophysical principles. This allows the user the 00002900 CU freedom to create the optimum processing sequence (not always 00003000 CU possible with package oriented systems). For example, the user 00003100 CU may select use of coherency statics (COST), surface consistent 00003200 CU statics (TRAC) or combinations of both (TRAC - COST). 00003300 CU 00003400 CU In a single SPARC job the user may select more than one sequence 00003500 CU of processes for the same data. For example, one sequence may 00003600 CU include stack and filter, another deconvolution, stack, and filter,00003700 CU and a third stack, deconvolution, and filter. This facility of 00003800 CU generating multiple outputs with different processing in each, 00003900 CU reduces the number of jobs required for evaluating various 00004000 CU processing alternatives. 00004100 CU 00004200 CU SPARC processes fall into two main categories. First, most SPARC 00004300 CU processes handle seismic data, each performing some specific 00004400 CU action such as NMO or filtering. 00004500 CU 00004600 CU A second category of processes handle parameter information such 00004700 CU as recording geometry, velocity, surface elevation and statics. 00004800 CU Many of these processes produce printer or graphics plots for 00004900 CU quality control. 00005000 CU 00005100 CU SPARC provides several analysis tools to the user. Several 00005200 CU velocity analysis techniques are available. Noise tests can be 00005300 CU performed in the F-K domain. Also included are processes for 00005400 CU frequency content determination and wavelet estimation. 00005500 CU 00005600 CU New processes are added to SPARC on a continuing basis. This, 00005700 CU plus enhancements of existing processes, provides for a steady 00005800 CU evolution of SPARC. This evolution has kept SPARC current with 00005900 CU the rapidly changing technology of seismic data processing. 00006000 CUEND 00006100 CUPAGE BASICS CUINDX CUINDX SPARC BASICS CU Functional Characteristics CU CU SPARC has been implemented on two types of computers. The first CU type consists of large scale IBM computers. The following list CU shows some of the main components: CU CU CPU - 3090 with Vector Facility CU CU Memory - 128 megabytes or more CU CU Peripherals - numerous 9 track tape drives and disk CU drives, IBM 3480 tape cartridge drives, and electrostatic CU plotters. CU CU Uses IBM MVS/ESA operating system CU CU The main steps of a SPARC job on an IBM computer are shown below. CU CU 1. Job is submitted from TSO. CU 2. Computer selection and input data set verification. CU This step is called JOBGEN. CU 3. Extensive parameter checking and table generation (used CU in step 4). This step is called PREP. CU 4. Process seismic data. This step is called PROC. CU CU SPARC characteristics particular to the IBM environment: CU CU 1. Automated accounting of processing costs. CU 2. Automated space management for tape and disk. CU 3. Supports running of large complex jobs. CU CU The second computer type consists of large scale CRAY CU supercomputers. The following list shows some of the main CU components: CU CU CPU's - Y/MP CU CU Memory - 64 megabyte (8 megaword) or more CU CU Peripherals - numerous cartridge tape drives and disk CU drives, plus access to IBM disk and electrostatic plotters. CU CU Uses CRAY operating system (UNICOS) CU CU The main steps of a SPARC job on a CRAY computer are shown below. CU CU 1. Job is submitted from TSO. CU 2. Computer selection and input data set verification. CU This step is called JOBGEN. CU 3. Submission of the job to the CRAY. CU 4. Extensive parameter checking and table generation (used CU in step 5) on the CRAY. This step is called PREP. CU 5. Process seismic data on the CRAY. This step is called CU PROC. CU 6. IBM job submitted back from the CRAY to update IBM data CU bases such as accounting and data set management. CU CU SPARC characteristics particular to the CRAY environment: CU CU 1. includes many processes, especially CPU intensive ones CU 2. automated accounting of processing costs. CU 3. automated space management for tape and disk. CU 4. supports running of large complex jobs. CU CUEND CU Coding Specifications CU CU An example of the seismic data cards needed to process a line CU is shown in Figure 1. This is a straight line with four skips CU in it. The output is to consist of five different stacked CU sections. CU CU This example shows the card formats. All but two of the cards CU are of a form that has nineteen data fields. The last fourteen CU fields (card columns 11-80) are five-column fields with the first CU five fields taking a different form. CU CU The first four card columns are defined as data field one. It is CU used to indicate the process name. The next column defines the CU process number which may be 0 through 9. This number allows each CU process to be used up to ten times with different parameters. A CU blank is interpreted as zero. The process name and process number CU uniquely define a set of parameters. This is considered a separate CU process from other processes with the same name but different CU process numbers. CU CU The third data field is also a single column. It designates CU the manner in which the parameters are to be spatially CU interpolated. The fourth data field is a single column and defines CU the process mode (file/shotpoint/depth-point/receiver). The next CU three columns, data field five, usually define a data card type. CU CU There are five seismic data cards that do not use this nineteen CU field format. One is the 'ACCT' card which chiefly contains CU accounting information. Another is the 'PROC' card, which is free CU format i.e. no 5 column fields. 'PROC' is used as an abbreviation CU for process. Both 'COMM' and 'HEAD' cards supply information used CU only for documentation purposes. The 'ULIB' card defines special CU program libraries to be used by the PREP and PROC steps. CU CU A specific example of a typical seismic data card can be taken CU from the example of Figure 1. 'FLTR2' defines a filter process CU as indicated by 'FLTR'. The process number, '2', distinguishes CU it from 'FLTR1' and 'FLTR3'. Linear interpolation of the CU parameters and depth-point processing are specified. The process CU mode range (of depth points in this case) is 176 to 344. Two CU data cards associated with 'FLTR2' are 'WIN' and 'TZF' cards CU which immediately follow the first 'FLTR2' card. These cards CU specify the design and application windows (WIN) plus the CU frequency response of the filter (TZF). CU CU EJECT CU CU Processing parameters that are interpolated should receive CU careful consideration. Data field 3, column 6, of the first card CU for each process specifies the spatial interpolation. At present CU the only options available are linear interpolation and no inter- CU polation. CU CU All traces in the range of shotpoints or depth points specified CU on a single data card will be processed using the parameters on CU that card. CU CU If a trace is between ranges specified on consecutive data cards, CU the processing depends on the interpolation. In the case of linear CU interpolation the trace will be processed with parameters inter- CU polated between those given on two consecutive data cards. If CU no interpolation is specified, the trace will not be processed CU at all. CU CU The 'MUTE1' cards in Figure 1 give an example of interpolation. CU The parameters on the first 'MUTE1' card will be used for the CU range of shotpoints on that card, namely 101 to 110. Similarly, CU the parameters on the second 'MUTE1' card will be used for CU shotpoints 150 to 159. The parameters for shotpoints 111 to 149 CU will be spatially interpolated between the values on the two CU 'MUTE1' cards. CU CU Column 6 of the first 'MUTE1' card is blank, so the interpolation CU defaults to that specified on the 'LINE' card. Column 6 of the CU 'LINE' card is also blank, so the interpolation defaults to 'L' CU for linear. CU CU Generally, the order of the seismic data cards is not important. CU The 'ACCT' card must precede the 'LINE' card within the deck. CU These two cards are required and so a suggested method of deck CU construction is to place these cards at the front of the CU seismic data cards. CU EJECT CU8080 ACCT 21598 0112345 SW74-17 SPARC SYSTEM TESTS 20SP 50 LINE S 101 159 59 48 4000 4 220 24 PROC GM3D 50 176 GM3D CLD 1 1 100 24 300 X 300 25 100 48 GM3D CFS 101 1 1 8 100 100 1 101 GM3D CFS 112 9 9 400 400 GM3D CFS 113 10 10 100 100 GM3D CFS 118 11 11 500 500 GM3D CFS 119 12 21 100 100 GM3D CFS 130 22 22 200 200 GM3D CFS 131 23 41 100 100 GM3D CFS 151 42 42 200 200 GM3D CFS 152 43 50 100 100 GM3D KIL 101 105 1 -26 GM3D KIL 106 106 1 -21 23 -26 GM3D KIL 107 107 1 -20 23 -26 GM3D KIL 108 108 1 -19 23 -26 GM3D KIL 112 1 -15 23 -26 GM3D KIL 113 1 -14 23 -26 GM3D KIL 118 1 -10 17 18 23 -26 GM3D KIL 119 1 -8 15 16 23 -26 GM3D KIL 120 1 -7 14 -16 23 -26 GM3D KIL 121 1 -6 8 23 -28 GM3D KIL 122 1 -5 7 8 23 -26 GM3D KIL 123 1 -5 23 -26 GM3D KIL 124 1 -3 23 -26 GM3D KIL 125 1 2 23 -26 GM3D KIL 126 1 23 -26 GM3D KIL 127 23 -26 GM3D KIL 128 4 23 -26 GM3D KIL 130 19 23 -26 GM3D KIL 131 133 23 -26 GM3D KIL 134 9 23 -26 GM3D KIL 135 3 23 -26 35 -38 GM3D KIL 136 2 23 -27 35 36 GM3D KIL 137 6 12 23 -26 34 -37 GM3D KIL 138 140 23 -26 GM3D KIL 141 3 23 -27 29 30 48 GM3D KIL 142 1 23 -26 29 30 47 48 GM3D KIL 143 1 22 -29 46 -48 GM3D KIL 144 1 22 -29 44 -48 GM3D KIL 145 17 22 -30 43 -48 GM3D KIL 146 1 10 15 22 -28 42 -48 GM3D KIL 147 1 14 22 -27 40 -48 GM3D KIL 148 6 13 19 21 -27 40 -48 GM3D DSP MAP DPS DIS VELF 101 5720 VELF TVR 101 0 5500 640 6000 840 6300 1280 6800 1760 7500 VELF TVR 101 2360 8400 400010000 Example of Seismic Cards Figure 1 CU EJECT READ1 F 1 50 14 STAT1 101 -3 5500 STAT1 SPE 101 -3 STAT1 SPE 112 -4 STAT1 SPE 119 -6 STAT1 SPE 124 -7 STAT1 SPE 138 -7 STAT1 SPE 144 -6 STAT1 SPE 151 -4 -3 -2 -1 0 1 2 3 FLTR1LS 101 159 101 FLTR1 WIN 101 1 FLTR1 TZF 1 7 13 37 43 NMOC1LS 101 159 101 0 2700 500 MUTE1 101 110 LFD 50 660 322 5720 1260 MUTE1 150 159 LFD 50 660 422 5720 1360 GATH1 176 344 STAK1 D 176 344 AFU WRIT1 D M DCON1LD 176 344 1 DCON1 TMF 1 128 36 10 600 3500 FLTR2LS 176 344 11 FLTR2 WIN 11 35 FLTR2 TZF 35 7 13 37 43 WRIT2 D M EQUL1LD 176 344 ABS 3000 0 4000 STAK2 D 176 344 AFU EQUL2LD 176 344 ABS 3000 700 2800 WRIT3 D M DCON2 D 176 344 2 DCON2 TMF 2 128 36 10 600 3500 FLTR3LS 176 344 2 FLTR3 WIN 2 2 FLTR3 TZF 2 7 13 37 43 WRIT4 D M AGCS1 D 176 344 RMS 3000 1000 WRIT5 D M PROC GM3D VELF READ1 PROC STAT1 FLTR1 NMOC1 MUTE1 GATH1 NODE1 PROC BRAN1 STAK1 NODE2 PROC BRAN2 WRIT1 PROC BRAN2 DCON1 FLTR2 WRIT2 PROC BRAN1 EQUL1 STAK2 EQUL2 NODE3 PROC BRAN3 WRIT3 PROC BRAN3 DCON2 FLTR3 NODE4 PROC BRAN4 WRIT4 PROC BRAN4 AGCS1 WRIT5 Example of Seismic Cards Figure 1 (Continued) CU8080 CUEND CU Processing Considerations CU CU The seismic data flow is essentially trace-sequential. A trace CU is read and then that trace is passed from one process to the CU next. Typically everything is done to the trace that is possible CU before the next trace is read. The only exception to a trace CU being completely processed is the inclusion of a multi-trace CU process in the processing sequence. For example, when the data CU is stacked, no traces are passed beyond this process until a CU stacked trace is completed. After the last trace in a gather CU is added to the stack, it is then passed to the following CU processes until it reaches another multi-trace process or else CU the process sequence is completed. This data flow has several CU benefits. The chief benefit is that the data is output as soon CU as possible so that a minimum of data is lost if a job aborts. CU Another benefit is that processes can be done in a logical order CU without concern for storing intermediate data. These benefits CU are provided automatically without the need for special parameters CU on the seismic cards. CU CU The seismic data processing sequence is determined by the 'PROC' CU cards. The processing sequences for both the PREP and PROC steps CU are completely determined by the order of processes on the 'PROC' CU card. There are some preparation processes that do not process CU seismic data; they only establish parameters. These processes CU should appear on the 'PROC' card(s) before the information would CU be needed. An example of this type of process is geometry or CU 'GM3D'. It does several things including the creation of the CU trace headers which almost every process must use. When 'GM3D' CU is needed, it will usually be the first process of the sequence. CU CU The data flow can be altered at any point in the processing CU sequence. This data flow change is effected by temporarily CU storing data, and then using this stored data at a later point CU in the processing sequence. The coding to achieve this change CU uses the symbols 'NODE' and 'BRAN' (indicating nodes and CU branches). These symbols appear only on the 'PROC' cards. CU Appended to these words is a digit, 0-9. the designation of a CU node indicates that data at this point is to be saved for later CU use. The digit differentiates this node from other nodes in the CU processing sequence. The branch defines the data to be used with CU the process in that branch. The digit associated with 'BRAN' CU refers to a previously defined and similarly numbered 'NODE'. CU The data flow continues in sequence until another 'BRAN' or 'NODE' CU is encountered. The sequence of 'NODE', 'BRAN' continues the CU process sequence as if 'NODE' and 'BRAN' had not been used, but CU allows the data saved at 'NODE' to be used at a later point in CU processing. There can be a maximum of ten nodes defined in a CU processing job. There may be any number of branches referring to CU a 'NODE'. Note again that the number associated with 'BRAN' CU EJECT CU CU refers to the node that has the desired intermediate data. In CU order to depict the data flow in each job, a summary page is CU printed at the completion of a job showing the relationship of CU all the nodes, branches, and processes. CU CU The use of 'BRAN' and 'NODE' is illustrated in the example CU of Figure 1. Figure 2 shows the flowchart of this job and CU will aid in understanding the coding of the 'PROC' cards. For CU example, consider 'NODE3'. The seismic data stored at 'NODE3' CU in the example has been gathered, equalized, stacked, and CU equalized after stack. One branch that refers back to 'NODE3' CU simply outputs this data. Another branch, however, uses this CU data by supplying it to 'DCON2' followed by 'FLTR3'. This CU branch is then broken by 'NODE4'. This example job demonstrates CU multiple nodes and multiple branches. 'NODE' and 'BRAN' are CU used on the 'PROC' card only if they are needed. CU CU The coding of the 'PROC' cards in the example demonstrates only CU one method. Many of the 'PROC' cards begin with 'BRAN'. This is CU not required, although most users find it convenient. The only CU rules are: CU CU 1. multiple 'PROC' cards must be in the correct order CU in the deck; CU CU 2. blanks or commas must separate the names on the cards; CU CU 3. omitted process numbers (including those for 'NODE' and CU 'BRAN') are assumed to be zero; CU CU 4. all branches for a node must be completed before a branch CU from an earlier node is started. CU CU 'GM3D' ('GEOM') must be used by the system unless only depth CU points in complete ARCY format are being processed. This CU typically means that the data has been previously processed CU through SPARC. If depth points are being read, but any process is CU in shotpoint mode, then 'GM3D' ('GEOM') is necessary. CU CU To insure that condition (4) above is met when the processing CU sequence includes a complex set of branches and nodes, it is CU convenient to prepare the 'PROC' cards so that they resemble a CU flowchart. CU EJECT CU CU As an illustration, the 'PROC' cards for the example in Figure 1 CU can be written as follows: CU CU PROC GM3D0 VELF0 READ1 STAT1 FLTR1 NMOC1 MUTE1 GATH1 CU PROC NODE1 CU PROC BRAN1 STAK1 CU PROC NODE2 CU PROC BRAN2 WRIT1 CU PROC BRAN2 DCON1 FLTR2 WRIT2 CU PROC BRAN1 EQUL1 STAK2 EQUL2 CU PROC NODE3 CU PROC BRAN3 WRIT3 CU PROC BRAN3 DCON2 FLTR3 CU PROC NODE4 CU PROC BRAN4 WRIT4 CU PROC BRAN4 AGCS1 WRIT5 CU CU CU Since the order of the branches for a node is not important, these CU 'PROC' cards could also be arranged in the following order: CU CU PROC GM3D1 VELF0 READ1 STAT1 FLTR1 NMOC1 MUTE1 GATH1 CU PROC NODE1 CU PROC BRAN1 EQUL1 STAK2 EQUL2 CU PROC NODE3 CU PROC BRAN3 DCON2 FLTR3 CU PROC NODE4 CU PROC BRAN4 AGCS1 WRIT5 CU PROC BRAN4 WRIT4 CU PROC BRAN3 WRIT3 CU PROC BRAN1 STAK1 CU PROC NODE2 CU PROC BRAN2 DCON1 FLTR2 WRIT2 CU PROC BRAN2 WRIT1 CU CU CU These arrangements make it plain that there are two 'BRAN1' CU branches attached to 'NODE1'. Either 'BRAN1' may be started CU first, but it must then be completed (with all its subsidiary CU nodes and branches) before the other 'BRAN1' is started. CU EJECT CU CU GM3D0 CU CU VELF0 CU CU READ1 CU CU STAT1 CU CU FLTR1 CU CU NMOC1 CU CU MUTE1 CU CU GATH1 CU CU NODE1 ----------------------- CU | | CU BRAN1 BRAN1 CU CU STAK1 EQUL1 CU CU NODE2 ---------- STAK2 CU | | CU BRAN2 BRAN2 EQUL2 CU CU WRIT1 DCON1 NODE3 ---------- CU | | CU FLTR2 BRAN3 BRAN3 CU CU WRIT2 WRIT3 DCON2 CU CU FLTR3 CU CU NODE4 ---------- CU | | CU BRAN4 BRAN4 CU CU WRIT4 AGCS1 CU CU WRIT5 CU CU CU CU Flowchart of Processes for Job of Figure 1 CU Figure 2 CUEND CU Job Submission and Procedures CU CU After the seismic data cards have been prepared, other cards are CU still required to execute JOBGEN for an IBM or CRAY job. CU Figure 3 shows the normal deck structure for EDP's geophysicists CU and Figure 4 shows the deck structure for District submission of CU SPARC jobs. Figure 5 gives an overview of the SPARC system. CU CU For an IBM or CRAY job from IBM TSO: CU CU 1. the deck must have the IBM job card(s) first. CU 2. the job name is seven characters specified as the TSO user CU id and a one digit job number. CU 3. the xxxxx represents the Accounting Organization Code and CU is unique for each different organization. CU 4. the four digits after xxxxx are the four digit CU project number (for project numbers > 10000 use the last CU four digits). CU 5. the user name should be coded last name followed by CU initials. It should be 5 to 20 characters. CU 6. the 'EXEC' card specifies the execution of SPARC. CU 7. the card following the 'EXEC' is also an IBM card that CU indicates the seismic data cards are to follow. CU 8. column 12 of the SPARC 'ACCT' data card indicates to which CU system (IBM or CRAY) the job is to be routed. CU If column 12 is left blank, the job will default to the IBM CU unless the job contains a CRAY only process. If a CU mixture of computer dependent processes occurs, the job CU will abort with an error message. CU 9. the last card is an IBM end card. This card is optional. CU CU After the job deck has been prepared the next events are CU straightforward. The first time a new deck is used it will CU probably be a 'PREP' run. A PREP run is specified by the CU characters 'PREP' in columns 77-80 of the LINE card. This will CU enable the parameters to be verified before processing the data. CU The prepared job deck is submitted from TSO. When the job is CU executed, the JOBGEN output will be available very quickly. If no CU error has been detected, JOBGEN automatically submits the second CU job to the IBM. The printer output of the second job is then CU the one that is of the most interest. The second job name for CU an IBM job consists of the seven characters of the first job name CU plus a one-character suffix. The suffix is 'A' for a PREP job and CU 'B' for a PROC job. CU CU EJECT CU If the geophysicist wishes to check his processing parameters CU without submitting a background job, he may do so by typing "PREP" CU on the TSO command line (at the top of the screen in EDIT (PDF 2)). CU To verify parameters for a single process type "PREP" followed by CU the process name e.g. PREP FLTR1. Note that the process number CU should be included. A message will be written to the screen CU indicating whether or not errors have been detected. The actual CU error messages may be viewed on the HOLD queue (PDF 8.h). Select CU the jobname which is the user's TSO ID. CU CU When the geophysicist is satisfied that the parameters are correct CU a 'PROC' run can be made. The only change that is needed in the CU card deck is to replace 'PREP' with 'PROC' on the 'LINE' card. To CU run on the IBM or CRAY, the trace-sequential seismic data must be CU at the EDP data center on tape or disk. The deck for the CU processing run is submitted for execution in the same way as the CU PREP job. If seismic data is output by the processing job, it CU is recorded on the printout and is available at the EDP data CU center. Each SPARC output process contains a parameter to request CU automatic plotting of the data. Refer to EDP GEOPLOT User CU Documentation for information on plotting. CU CU EJECT CU EDP JCL CU *** *** CU CU //tsouidn JOB (xxxxxphon,,,tpanum),'username,i.i.',CLASS=U | CU ------- --------- ------ ------------- CU /*USER S=SEISMIC CU /*USER G=JOBGEN CU /*ROUTE PRINT LOCAL CU //stepname EXEC SPARC | CU -------- CU //SEIS.CARDS.DD * CU ACCT CU LINE CU . CU . seismic data cards CU . CU PROC CU // CU CU CU Underlined items are variable. (Do not enter underlines.) CU CU tsouid is your TSO user id and n is any one digit number. CU CU xxxxx Accounting Organization Code, GAP = 45373. All CU other departments have their own unique code. CU CU phon phone extension, four digits. CU CU tpanum TPA number, six digits. CU CU username,i.i. user's name and initials, 5 to 20 characters. CU CU stepname 1 to 8 characters, first is alphabetic. The CU stepname is optional. CU CU CU IBM and CRAY SPARC TSO Deck Structure CU Figure 3 CU EJECT CU CU CU District JCL CU ******** *** CU CU //tsouidn JOB xxxxx95ww,'username,i.i.',CLASS=U | CU ------- ----- -- ------------- CU /*USER S=SEISMIC CU /*USER G=JOBGEN CU /*ROUTE PRINT N1Ryyy CU --- CU //stepname EXEC SPARC | CU -------- CU //SEIS.CARDS DD DATA,DLM=$$ CU ACCT CU LINE CU . CU . seismic data cards CU . CU PROC CU /*ROUTE PRINT N1Ryyy CU --- CU /*JOBPARM ROOM=zzzz CU ---- CU $$ CU CU CU Underlined items are variable. (Do not enter underlines.) CU CU tsouid is your TSO user id and n is any one digit number. CU CU xxxxx Accounting Organization Code. Use the unique CU code assigned to your group. CU CU ww district number (see ACCT card) CU CU username,i.i. user's name and initials, maximum 20 characterS. CU CU yyy remote sysout location. CU CU stepname eight characters or less, first is alphabetic. CU CU zzzz room number. CU CU CU CU CU IBM and CRAY SPARC District TSO Deck Structure CU Figure 4 CU EJECT CU CU JOBGEN Options CU CU CU The default output for a successful JOBGEN will include only a CU "TERMINATED NORMALLY" message and an Output Dataset Report. A job CU with errors will print appropriate error messages. To override the CU default listings, the following keywords may be entered in the CU JGPARM parameter on the EXEC card: CU | CU CARDS - prints only the input cards. CU DUMMPROC - use dummy process for any unknown processes. This CU is the default option for non-EDP SPARC jobs. If CU not specified for EDP SPARC jobs, any unknown CU process will cause an error message. CU SUMMARY - the data cards and PROC summary will automatically CU be printed for the job. CU SYSOUTA - use MSGCLASS=A for SPARC job. CU SYSOUTH - SPARC job will be put in the output hold queue. CU NOEXTEND - do not use extended addressing on IBM. CU CU You may specify any combination of the above keywords. For example CU CU // EXEC SPARC,JGPARM='CARDS,SYSOUTH' CU CU will generate listings of the input cards and put the SPARC job in CU the output hold queue. CU CU JOBGEN will default the MSGCLASS of the submitted SPARC job based CU upon whether it is a PREP or PROC job. The defaults are: CU CU PREP jobs - MSGCLASS=H CU PROC jobs - MSGCLASS=H CU CU The 'A' MSGCLASS means that the job is immediately available for CU printing on a line printer. The 'H' MSGCLASS means that all print CU is in a 'HOLD' status and may be viewed through 'PDF 8.H' by CU placing an 'S' under 'NP' of the line containing the job name. CU After reviewing the output, you may delete the output by placing a CU 'P' under 'NP' of the line containing the job name or the output CU may be printed by placing an 'O' under 'NP' and an 'A' under 'C' on CU the line containing the jobname. Please delete all the outputs not CU needed as soon as you can, in order to conserve system disk space. CU CU The default MSGCLASS for any type of job may be overridden by using CU keywords 'SYSOUTA' or 'SYSOUTH' in JGPARM field. These keywords CU may be used with the other keywords described above. CU CU Instead of waiting for the printed output from JOBGEN, you may use CU the 'PDF 8.H' options already described to display the JOBGEN CU output on your crt screen. To use this facility add the CU EJECT CU CU 'MSGCLASS=H' parameter to your JOBGEN job card as shown below. CU CU //tsouidn JOB (45373proj,,,tpanum),'username,i.i.',CLASS=J, | CU ------- --------- ------ ------------- CU // NOTIFY=tsouid,MSGCLASS=H CU ------ CU All other jcl is the same as in Figure 3. CU CU CU NOTE: JOBGEN and SPARC output that is placed on the "Hold Queue" CU (MSGCLASS=H) as shown above will be automatically deleted after the CU hold queue retention period (currently two days). If you need a CU permanent copy, you must print it before it is deleted. CU EJECT CU CU ----------- CU | SEISMIC | CU | DATA | CU | CARDS | CU ------------ CU | CU v CU -------------- ------------- CU | | | | CU | JOBGEN |-------->| LISTINGS | CU | | | | CU -------------- ------------- CU | CU v CU ************** CU * GENERATED * CU * SPARC JOB * CU *(INTERNAL TO* CU * IBM) * CU ************** CU | CU v CU ------------- -------------- ------------- CU | DEMUXED | | | | OUTPUT | CU | SEISMIC |-------->| SPARC |-------->| SEISMIC | CU | DATA TAPE | | | | DATA | CU ------------- -------------- ------------- CU | | CU v | CU -------------- | CU | SPARC | v CU | PRINTED | ----------- CU | LISTINGS | | DISPLAY | CU -------------- | SEISMIC | CU | DATA | CU ----------- CU CU CU SPARC System Overview CU Figure 5 CUEND CU System and SPARC Completion Codes CU CU Completion codes are issued by both the IBM control and SPARC CU application programs to indicate when a task or job has abnormally CU terminated. Both the system and application completion codes CU reflect important information and should be investigated fully CU before any action is taken. Problem determination usually CU requires the help of either operation or programming support. CU CU CU I. IBM System Completion Codes CU CU System completion codes or abends indicate that the control CU program has determined that a job cannot continue processing CU and expect valid results. For example, the task may have attempted CU an invalid operation, or an input/output operation may have CU failed. In such a case, the system terminates the job and supplies CU the completion code to indicate the reason for the termination. CU The code may be any three-digit hexadecimal number and will appear CU in the JES2 joblog section of the job printout. CU CU All the system completion codes are listed in the IBM OS/VS MESSAGE CU LIBRARY publication; however, to avoid any erroneous problem CU determination and/or any subsequent inappropriate action, the CU system codes are not readily available to the SPARC users. These CU system messages must be monitored by programming or operations CU support on a regular basis to isolate reoccuring hardware and CU system related problems. CU CU A few examples of system codes and actions are as follow: CU CU code explanation action CU ---- ----------- ------ CU CU S001 hardware I/O error inform programming support CU immediately. CU CU S222 operator canceled job contact operations support CU for cancellation explanation. CU CU S322 specified job time limit break up job or specify a CU exceeded larger cpu time and resubmit CU job. CU CU CU These are only three of the more frequent system completion codes. CU They, like all the others, have their own unique definition and CU subsequent action. These codes aid in problem determination and CU are a necessity in identifying failing hardware or programs. CU EJECT CU Contrary to IBM standards, actions should not be taken to resolve CU an abnormal job completion without full knowledge of the problem CU at hand. CU CU CU II. SPARC Application Completion Codes CU CU SPARC application completion codes and diagnostics are an CU indication that the problem or process has determined that its CU parameters or processing are in error and that it cannot continue CU processing expecting valid results. Unlike system codes, the CU application codes and diagnostics are usually meaningful only for CU the specific process in error and therefore cannot be explained CU here. They should, however, be totally self explanatory and CU subsequent user action should be very straight forward. CU CU Sometimes application codes and messages are not very clear and CU problem determination may require the help of programming support. CU Unlike system codes, SPARC application abends are usually fatal and CU the jobs cannot be resubmitted without some definite action. The CU user can tell very quickly if a given SPARC job has abandoned via CU the application program by investigating the PREP/ PROC process CU flowchart summary at the end of the job. An '(A)' will be printed CU after all the processes in the flowchart that have abandoned and CU the user can turn to the respective process sections in the CU printout for more information about the abnormal completion. CU CU For both system and application abnormal job completions, the user CU should make every attempt to salvage any processing prior to the CU abend. Abnormal completions do not imply that the entire job is CU unusable. Patching a job rather than a total resubmit can CU eliminate much duplication of effort and computer resources can be CU freed providing for a much more effecient processing system. The CU user should see programming support any time help is required to CU patch a job after partial job completion. CUEND CU Glossary CU CU ARCY (ARC-Y) - trace header format similar to SEG-Y but with CU ARCO additions. CU CU D - depth point. CU CU DF - definition field or data field. Used to CU number the parameters on a SPARC data card. CU CU DATA SET - a collection of data that is organized to be CU used by the computer. Typically the storage CU medium is tape, disk, or mass storage. CU CU DEPTH POINT - numbers that are generated from the geometry CU NUMBERS information in ascending, sequential order. CU Thus, these numbers are spatially meaningful. CU These numbers are stored in the trace headers. CU CU DISTANCE - the offset measured from the shot in units of CU distance (usually feet or meters). CU CU F - file. CU CU FILE NUMBER - a name given to a set of traces that are CU grouped together. This number is typically CU a sequence number that has been generated CU outside of SPARC. The number, however, is CU retained in the trace header. CU CU GI - geophone group interval. CU CU JOBGEN - this is the name of the program that is CU executed to submit the seismic job. CU CU L - linear. CU CU MS - millisecond. CU CU NO. - number. CU CU NUMERIC ID - this an identification number. It is used to CU refer to other cards within the same process. CU It is recommended that this number be chosen CU to refer to a shotpoint or depth point and CU thus give it meaning associated with the line. CU CU O - offset. CU EJECT CU CU OPT - optional. This usually indicates parameters CU that are coded at the discretion of the person CU processing the data. CU CU PROCESS NUMBER - This optional number allows the specifi- CU cation of processes with the same function CU but different parameters. A blank is CU interpreted as zero. CU CU PROCESSING MODE - code in DF4 of process card to indicate type CU of data to be processed: files, shotpoints, CU depth points, receivers, or offsets. CU CU PROCESSING MODE - refers to the numbers coded in DF6 and DF7. CU RANGE Either shotpoint numbers, depth point numbers, CU file numbers, receiver numbers, or offsets. CU CU R - receiver. CU CU REQ - required. Indicates fields that must be coded CU on the seismic data cards. CU CU S - shotpoint. CU CU SHOTPOINT NUMBER - a number assigned to a group of traces. Within CU SPARC this number has no spatial or geometrical CU significance. A number should be used only CU once on a seismic line. CU CU SI - sample interval. Units used are milliseconds CU unless otherwise stated. CU CU VSP - vertical seismic profile. CUEND