RFP_QA.docx

Questions and Answers

1. GOJCL has references to COBOL COPYLIB YBPST4.A140.BENCH.COPYLIB. Please provide a copy of the library or indicate that it is not required.

A copy library is not required. Each of the four source code files is self-contained. We reference a copylib as a matter of convenience.

2. File Transfer Benchmark - Are there execution results of the File Transfer Benchmark that can be provided?

We cannot provide precise execution results at this time.

3. For the XPBM benchmark - Are there a number and/or time of runs that are required?

The XPBM Benchmark should be run a minimum of three times to establish the billing reference. For our evaluation, we will use the average of the three runs shown in the Benchmark Instructions, which is 41.98 CPU seconds (IBM z10 2097-402). We expect a relatively small variance among iterations.

4. What is the Government's anticipated/ideal transition period for this effort?

The offeror should propose the transition period, as it is part of the evaluation. Please reference page E-5 of the RFP.

5. Please provide the formula or describe how the Equivalency Factor of 40.70 was derived for the IBM z10 2097-402.

This was established with an earlier benchmark suite. The suite was initially run on an IBM 200E processor and was rerun whenever a processor upgrade was put in place. The ratio of processing times on the new CPU, compared to the 200E run, was calculated and became the new Equivalency Factor.

6. Please provide the benchmark results for the 3090-200E which are being used in the baseline calculations for generating the Equivalency Factor.

We plan on comparing the execution times offerors indicate from their running of the benchmark to execution times in the incumbent environment. The ratios of these values, combined with the offered prices, will allow us to compare cost per unit of processing power. We can’t rerun the 2014 benchmark on a 200E environment as we don’t have access to one.

7. Page B1 indicates that the RFP respondent shall calculate and provide the Equivalency Factor. Page E11 implies that BLS will simply use our Benchmark results to calculate the Equivalency Factor. Please clarify.

See answer to question 6 above. For contract award purposes, we will use the offerors’ indicated benchmark results (in conjunction with their offered prices) to make our selection. We will rerun the suite once the conversion is complete to confirm/establish the factors to be used for billing. We would not expect a significant discrepancy.

8. Please confirm that the unit prices that must be entered in the Section B Pricing should be the respondents “Equivalency Factor” adjusted unit prices for the IBM 3090-200E.

Yes, the unit prices must be adjust unit prices for the IBM 3090-200E. The adjustment to IBM 3090-200E is for evaluation purposes. At time of contract award, we will update pricing to reflect the winning offeror’s native CPU seconds.

9. What was the total dollar amount spent against the BLS portion of the current contract for FY2013? What is the total dollar amount spent against the BLS portion of the current contract for FY 2014 year-to-date.

FY 2014: Oct 2013 – Feb 2014 = $507,406 FY 2013: Oct 2012 – Sep 2013 = $1,340,445

10. What was the total number of 3090 or Z10 CPU seconds used for the BLS portion of the current contract for FY2013? Please provide a breakdown by Batch Priorities 1 through 9, Interactive Processing Primetime, Interactive Processing Nights, and Started Tasks - CICS, VPS. Please provide the same for FY2014 to year-to-date.

See the spreadsheet that is attachment 4. It shows counts and CPU seconds on a daily basis for calendar year 2013 for the largest user account on the contract (about 90 percent of the total resource consumption). There are columns for the various batch job priorities, FTP transfers, the CICS regions, and TSO sessions. There is also data on the number of and page volume of LPS print jobs. When we compiled this data, we did not consider the difference between interactive processing in prime time and at night to be significant – the vast majority would be prime time. We also did not separately tally cycles and billing used by the VPS system; that data was not in the billing file these data were assembled from. Note that the CPU seconds in this table are in thousandths of a native CPU second. The processor was upgraded from a 2064-102 used in the January 1 to February 2 period to a 2097-402 for February 3rd forward.

11. Is BLS satisfied that a dedicated 50 MIP environment will meet their daily production requirements?

Yes, but we are not experts in the appropriate MIP environment needed to meet our daily production requirements. We certainly get good service from the 50 MIP environment at the moment. There is a significant possibility that a smaller environment might also meet our needs, but we don’t have the expertise to assess that. If it would be advantageous to a vendor to move us to a smaller environment without a significant impact on performance, BLS would consider approving such a move.

12. Approximately how big is the current Public Storage DASD pool?

Please reference page C-4 of the RFP where we describe our requirements for Public DASD.

13. Does any contractor LPS printed output require same day delivery in Washington DC?

No. Normally, none of the LPS output is sent to Washington DC (it goes to the state agencies). Furthermore, we don’t require next day printing, much less delivery, on any of the LPS output (testing would be the exception). We would expect to have printing and shipping occur within one week of the request (print job execution) unless it was a particularly large job, or a large coincident volume, in which case further delay might be acceptable.

14. CA-DADS is no longer commercially available. Will BLS provide the software or is an alternate product such as IBM’s Batch Application Control acceptable?

The "DADS" software is something that we had been unfamiliar with until we got into the SunGard system. We had used only CEMT and CEDA previously. These alternate products are acceptable. Alternative software that provides equivalent functionality are acceptable as well.

15. What specific software components of File-AID and Abend-AID product suites are required?

We are not really certain as to which of the specific File-AID components are needed or are not. We frequently use File-AID for viewing file contents (both VSAM and "flat" files), but occasionally we will use the edit mode as well. We have used the 3.6 option quite a bit to find all occurrences of a particular string in the members of a PDS. It is helpful to use the 3.5 option to view the characteristics of a VSAM file so that we know if it is getting overly fragmented. These are the parts of File-AID that we have been using.

Regarding Abend-AID: we need the interpreter portion of the dump that tells us what kind of error we had and what it means. We have not been using other features of Abend-AID.

16. Will BLS make available for migration full volume backups of all system, application, and user data?

We will make backups of application and user data available. We expect some will be available as full volume backups, some otherwise. The incumbent provider has an obligation to cooperate with the transition.

17. Regarding the subject solicitation, what is the anticipated award date and anticipated period of performance for this opportunity?

The anticipated award date is late spring/early summer

18. Who is the current incumbent providing these services?

Sungard Data Systems Inc.

19. When does the current contract end?

May 31, 2014.