Attachment_B_-_BLS_Mainframe_Benchmark_Suite_Instructions_-_June_2019.docx

The BLS File Mainframe Benchmark Suite January 16, 2019 The BLS Mainframe Benchmark Suite is composed of two parts, the XPBM Benchmark and the File Transfer Benchmark. The File Transfer Benchmark uses the same data files as provided for the XPBM Benchmark.

Please note that this Benchmark and the Instructions were compiled for a previous acquisition but that the procedures are still valid for this competition. The sample times contained here were compiled on equipment that BLS no longer uses and that the equivalency factor calculated is therefore out of date. Nevertheless, as examples of the process, these are sufficient to perform the Benchmark.

The BLS XPBM Benchmark February 28, 2014

The BLS XPBM Benchmark is a set of programs and data intended to be part of the Bureau of Labor Statistics’ Mainframe Service Center Benchmark Suite.

There are a total of 8 files that are necessary to run the benchmark.

There are two sequential data files:

prefix.MICRO is a 743,280,900 byte file;

prefix.LOOKUP is a 1,051,380 byte file.

There are 4 COBOL source programs:

prefix.MI28 is a 10,201 line file;

prefix.UT01 is a 1,712 line file;

prefix.UT19 is a 10,756 line file;

prefix.UT23 is a 7,294 line file.

There are two JCL streams:

prefix.COMP4 is a 39 line stream that compiles the 4 COBOL programs;

prefix.GOJCL is a 296 line stream that executes 16 steps including 4 IDCAMS, 6 IEFBR14s, 2 sorts, and each of the 4 compiled COBOL programs.

In transferring the sequential data files up to the MAINFRAME, those data files should have a binary transfer. The other files should be transferred using ASCII (to EBCDIC) conversion.

The source files had the following attributes before they were downloaded:

NOTE: Due to the large amount of data contained in the MICRO file, it and the others are provided in Zipped form.

Offerers will need to change the JCL to conform to their site’s standards.

In the provided jobstreams, the userid is YBPST4, the account is A14000000, the prefix for all files is YBPST4.A140. and for most it is the further qualified prefix of YBPST4.A140.XPBM. One facility specific storage volume cited is BLSWK2 – change as appropriate. One facility specific unit types is WRKANY – change as appropriate. Other changes may need to be made as well. Please provide a description of all changes and a copy of the modified jobstreams.

In addition to the source files, data, and JCL provided, sample output is provided.

The names, attributes, and sizes of the output are synopsized in the following table.

After running the jobs, the Offerer shall provide BLS with all output datafiles so that they can be compared electronically to the reference files. Printed/paper versions of all of the listings except the large Step15 listing – ddname ES2EXCPR – should be generated and provided. The first 10 pages of the excepted listing (ES2EXCPR) should also be generated and provided.

Aside from the time and date stamps, and any installation and version specific differences in the print output of the SORT and IDCAMS utilities, the listings should be identical to the sample listings.

If you have any questions or concerns, please contact the Contracting Officer.

The following table shows step execution times for each step of the XPBM as run by BLS in February 2014. Samp1 was run in the afternoon under priority class 4, and Samp2 and Samp3 were run on the weekend under priority class 7. The next page shows additional statistics for each of these runs. Step and job times are stated in unadjusted IBM z10 2097-402 CPU seconds. Note that the CPU seconds listed in the Attachment 1 – Pricing Tables are in terms of the IBM 3090-200E. The equivalency factor for the IBM z10 2097-402 is 40.7.

* JOB TERMINATION STATISTICS *

* JOB NAME YBPST4BM JOB75760 USERID YBPST4 CPU TIME 43.09 SRB TIME 1.12 *

* PROGRAMMER ES2J028D START DATE 02/07/14 14.038 START TIME 14:59:17.81 *

* SYSTEM ID CCF1 STOP DATE 02/07/14 14.038 STOP TIME 15:02:48.54 *

* CLASS 4 ELAPSED TIME 3:30.73 *

* VIO EXCP TAPE CONN TIME DISK/TAPE EXCP CPU CLOCK MOUNTS/DRIVES COST 200-E SECS *

* 0 .00 98,330/ 0 43.09 210.56 0 / 1 XXXXXX 1753.76 *

* JOB TERMINATION STATISTICS *

* JOB NAME YBPST4BM JOB81440 USERID YBPST4 CPU TIME 41.44 SRB TIME 1.02 *

* PROGRAMMER ES2J028D START DATE 02/09/14 14.040 START TIME 17:29:16.65 *

* SYSTEM ID CCF1 STOP DATE 02/09/14 14.040 STOP TIME 17:31:14.38 *

* CLASS 7 ELAPSED TIME 1:57.73 *

* VIO EXCP TAPE CONN TIME DISK/TAPE EXCP CPU CLOCK MOUNTS/DRIVES COST 200-E SECS *

* 0 .00 98,378/ 0 41.44 117.60 0 / 1 XXXXXX 1686.61 *

* JOB TERMINATION STATISTICS *

* JOB NAME YBPST4BM JOB81540 USERID YBPST4 CPU TIME 41.40 SRB TIME 1.02 *

* PROGRAMMER ES2J028D START DATE 02/09/14 14.040 START TIME 19:07:15.29 *

* SYSTEM ID CCF1 STOP DATE 02/09/14 14.040 STOP TIME 19:09:01.36 *

* CLASS 7 ELAPSED TIME 1:46.07 *

* VIO EXCP TAPE CONN TIME DISK/TAPE EXCP CPU CLOCK MOUNTS/DRIVES COST 200-E SECS *

* 0 .00 98,383/ 0 41.40 105.96 0 / 0 XXXXXX 1684.98 *

The BLS File Transfer Benchmark February 28, 2014 The BLS File Transfer Benchmark is part of the BLS’ Mainframe Service Center Benchmark Suite. All benchmark testing that the Offerer reports must be performed in compliance with the requirements of the Performance Work Statement regarding data transmission and security.

The File Transfer Benchmark consists of two files that are to be uploaded to the mainframe and downloaded from the mainframe. The two files to be used are the same two data files as used in the BLS XPBM Benchmark, known as MICRO and LOOKUP. File attributes for these files are specified in the XPBM Benchmark.

The Offerer may choose the configuration of the non-mainframe side of the file transfer. Any equipment or software that is used to perform this test that is not part of the service being offered to the Government shall be described in the Offerer’s response.

To conduct the File Transfer Benchmark, the MICRO file should be uploaded to the mainframe and then downloaded from the mainframe. Then in a sequence of five times, the LOOKUP file should be uploaded and downloaded from the mainframe. The table below shows what should be reported for this sequence. Fixed monthly or annual charges should not be reported in this table. It is the unzipped versions of these files that are to be transferred.

If an optional file transfer system has been proposed, an additional table should be included reflecting the performance and billing unit costs of that system.

Regardless of the file transfer method proposed, the Offerer shall show all billable resources pertaining to the transfer in the tables below. More than one table may be submitted, if necessary.

If a proposed FTP method transfers compressed files, any marginal costs of compressing and expanding the files on the mainframe should be included in the table. BLS assumes the marginal cost of zipping (compressing) and unzipping (uncompressing) downloaded versions of the files to be negligible. In the case of proposing the transfer of compressed files, the Offerer should provide digital copies of the compressed files before and after transmission so that their integrity can be validated.