JPH0445861B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is directed to DASD (Direct Access Storage) of fixed length block architecture (FBA).
Device) has a partitioned data set (divided into a directory part and a member part area, the directory part has information indicating the member name and the location where the body exists, and the member part has a structure that contains the member body. For the purpose of converting a data set (a data set) into a partitioned data set that can be used in a variable length (CKD) format data system with different data lengths that can be accessed at one time.
This is a method to convert a partitioned data set on DASD to a JSECOPY unloaded version.

There are two main types of data formats in DASD: fixed-length block data and variable-length data.
There are different types, each with a different address expression.

In other words, a partitioned data set on a DASD in FBA format is stored in a CI (Control
It consists of a collection of data called Intervals, all of which have the same length. This CI is composed of variable length blocks A, B, etc., and the position of the block is expressed by a 3-byte length CCB. Here CC is the relative CI number (0 to n) within the dataset,
B is the block number (0 to m) within each CI.
For example, block E in FIG. 2 is expressed in the form "000102" (hexadecimal) with CC=1 and B=2.
By the way, this data set is provided in file F as shown in Fig. 3, but the VTOC indicates where the start address of the data set with data set names α, β, γ, etc. is located.
(Volume Table Of Contents) is provided, in which data set names α, β, γ, . . . and addresses a, b, c, . . . indicating their starting positions are written as absolute addresses. And for each dataset,
As shown in Figure 4, a directory area in which member names G, H... and their relative addresses (described in the CCB) are written in alphabetical order, for example.
DRCT and a member area MNMB in which members G, H, . . . are actually written are provided. Therefore, when accessing member P existing in data set α, first access the VTOC in Figure 3.
Starting position a of data set α (absolute address)
Then, the member P is read from the directory area DRCT of the data set α, and the member P is written therein. The first address of block E (written in CCB) is obtained, and member P can be accessed from this.

In addition, in the CKD format DASD, the block position is expressed as a 5-byte long CCHHR. CC here
is a 2-byte long cylinder number, HH is a 2-byte long track number, and R is a 1-byte long block number. These relationships are as illustrated in FIG.

[Problems to be solved by the invention]

By the way, load modules created in the FBA system cannot be used in the CKD system, which has a different address representation. Therefore, in the past, it was necessary to bring the source program to the target system, compile it, and then combine and edit it to create a load module.
This is useful not only when migrating FBA's partitioned dataset to the CKD system's load module, but also in the reverse case, by bringing the source to each system, compiling, and combining editing to create the required partitioned dataset. I had to.

[Means for solving problems]

In order to solve the above problems, in the present invention, as shown in FIG.
It provides a migration tool to convert to an unloaded version of the program. A CKD classification data set can be obtained from the unloaded version created by this migration tool using the JSECOPY data set utility.

As a migration tool means for this purpose, the present invention uses a fixed-length format in which data in a fixed-length format is stored.
A DASD, a reading unit that reads the directory and member portion of a partitioned data set from the DASD, a unit that obtains the number of tracks per cylinder of the DASD from the operating system, and a start of the partitioned data set on the variable-length DASD. A virtual extent information generator consisting of an absolute address indicating the end, the number of tracks, and the number of extents, and a fixed length format from the operating system.
The part that acquires the CCB (*1) which is the address of the member part of the partitioned data set on the DASD, and the number of tracks per cylinder of the DASD (the CI number (CC part of the CCB) obtained from the above operating system). *2) An arithmetic unit that calculates the quotient and remainder, the cylinder number of the variable-length DASD that is the quotient, and the track number and block number of the variable-length DASD that are the remainder (part B of the CCB).
A concatenation part that concatenates the record numbers of a variable-length DASD that is used as is, and segmented data that is read from the DASD using the number of tracks per cylinder of the DASD obtained from the operating system and the virtual extent information as special information. The directory part of the set is used as a directory.
The cylinder number of the variable-length DASD, the track number of the variable-length DASD, and the record number of the variable-length DASD obtained from the above calculation results are added to each block of the member part of the partitioned data set read from the DASD, and the block is added to the external block. It is characterized by having a reading section that writes data to the data storage section.

*1: The fixed length forming DASD that is the input of this invention is
One type of dataset organization that can be stored in DASD in a format called FBA (Fixed Blocked Architecture) is an organization called partitioned dataset. The data in a dataset with this organization is a set of fixed-length data called CI, similar to a VSAM dataset. Furthermore, within the CI there is a collection of data called a block whose length is variable (see Figure 8).

Each CI is assigned a relative number from the beginning of the data set, and this is called a "CI number." Each block within a CI is assigned a relative number from the beginning of the CI. This is called a "block number."

*2: FBA is a fixed-length DASD, but its physical structure is as shown in Figure 5, and the number of tracks per cylinder can be obtained using operating system macros.

[Effect]

This allows you to easily obtain an unloaded version of the JSECOPY program needed to use FBA format data in a CDK format system.

〔Example〕

Before describing the present invention in detail based on one embodiment, an outline thereof will be explained.

For example, a partitioned dataset for FBA devices.
When used in a CKD system, it is necessary to convert to an unloaded version of the JSECOPY program.
This unloaded version is written on the magnetic tape MT in the format shown in FIG. This MT has
Special information area SA such as the number of tracks per cylinder, extent information in the data extent block (DEB), directory area DRA, member area MA where members A, B, etc. are entered.
1, MA2, etc. are stored. At the beginning of each block of this member data is the 8th
As shown in the figure, the disk address on the original partitioned data set (PO) is in CKD format.
Added by CCHHR.

By the way, the extent information in DEB is CKD
It consists of the number of divided areas (extents) of the data set, the start address, end address, number of tracks, etc. for each extent. For example, as shown in FIG. 7, data set A' includes track numbers 0 to 3 and 8 to
When the data is divided into 10 parts and stored, the number of extents is 2, and the first extent has a starting address of cylinder number 10, track number 0, and track number 4, which indicates its final address. Similarly, the start address of the second extent is track number 8, and the number of tracks is three. However, with this invention, this can be changed to JSECOPY.
When converting to the unloaded version of the program, the number of extents is 1, the data set start address is "00000000" (hexadecimal), the data set end address is "7FFF7FFF" (hexadecimal), and the number of tracks is "7FFF".
(hexadecimal).

The directory area DRA is as shown in Figure 4.
The first address where members are entered is stored. In FBA, this directory area is expressed as a relative address using CCB, but in the case of CKD, it is expressed as TTR. Here, "TT" is the relative track number within the data set, and "R" is the record number within the track. Therefore, the CCB of the relative address within the dataset is assumed to be the TTR within the dataset even in CKD. i.e. FBA DASD
Relative CI number of CKD DASD track number TT
Assuming that FBA's DASD block number B is
This is assumed to be the record number of CKD's DASD.

In addition, the MT data of the unloaded version of this JSECOPY program is shown in Figure 6.
CKD by JSECOPY dataset utility
The absolute address (CCHHR) required when writing to the DASD of the FBA is obtained from the relative address CCB of the FBA by the following calculation, and the absolute address CCHHR thus obtained is added to each member.

Next, we will explain how to calculate the absolute address (CCHHR) on the DASD from the relative address (CCB).

Operating system macros
DEVTYPE calculates the number of tracks TR per cylinder from the FBA DASD device information, and also uses the CCB (operating system macro)
(obtained by NOTE), calculate the relative CI number within the data set, that is, CC, and perform the calculation of the following equation (1).

CC/TR...(1) When the quotient of the above formula (1) is C ₀ C ₀ and the remainder is H ₀ H ₀ (both in hexadecimal), this C ₀ C ₀ is the cylinder after conversion to CKD. H ₀ H ₀ becomes the track number. Then, the CI block number B is set as R (record number).

The implementation of this conversion operation will be explained with reference to FIG.
In FIG. 8, when the number of tracks per cylinder of the FBA DASD is 10, blocks A and B are stored at relative address CC=1 in the data set,
Blocks C to E are stored at CC=2, block F is stored at CC=15, and blocks G and H are stored at CC=2.
16, and each block is stored in each CI as shown in Figure 8a.
Assume that they are stored in the order shown.

Block A has CC="0001", that is, 1, so
The calculation using the above equation (1) is 0001 (hexadecimal)/10=1/10, where the quotient is 0 and the remainder is 1. Also, B is "01". Therefore, the CCHHR format address after converting CKD to DASD is "0000000101" as shown in block A in Figure 8b, and this is added to the beginning of block A of member data in Figure 7b. .

Also, since block C has CC="0002", that is, 2, the calculation according to equation (1) is 0002/10=2/10, and the quotient is 0 and the remainder is 2. And since B is "01", the address in CCHHR format is
It becomes "0000000201".

Since block F has CC="000F", the calculation using equation (1) above is as follows: 000F/10=1...5 That is, the quotient is 1 and the remainder is 5. And since B is "01", the address in CCHHR format is
It becomes "0001000501".

And since block H is CC="0010",
The calculation using equation (1) is as follows: 0010/10=1...6 In other words, the quotient is 1 and the remainder is 6. Since B is "02", the CCHHR format address is "0001000602". In this way, each block in Figure 8a can be used for CKD as shown in Figure 8b.
Converted to CCHHR.

This completes the FBA DASD partitioned dataset.
Having described the method for converting to the JSECOPY unloaded version, we will also explain how JSECOPY uses this unloaded version to restore it to a partitioned data set for the CDK system, and demonstrate the validity of this migration tool.

The member part of the unloaded version of JSECOPY is the first
As shown in the figure, only the blocks with CCHHR added are arranged, and it is not possible to tell which block corresponds to which member just by looking at the member section. Therefore, first read the special information (including extent information) and directory part of the unload version. Then, from these two pieces of information, the beginning of the member is
Find CCHHR and find a block in the member section that has the same value as CCHHR. This situation is shown below.

In the case of CKD DASD, the address pointing to a member in the directory is expressed in TTR. "TT"
is the relative track number from the beginning of the data set, and "R" is the record number within the track.
Also, set the cylinder number of the extent start CCHH in the special information of the unload version to CCs, the track number to HHs, and the number of tracks per cylinder to n.
In this case, the absolute address CCHHR pointing to the beginning of the member can be found using the following formula. this
Correlate blocks of members with the same value as CCHHR by reading the member section of the unloaded version.

(TT+HHs)÷n ---→ | | | | ---→ Quotient + CCs → Cylinder number (CC) Remainder → Track number (HH) R → As is → Record number (R) Here, unload the data created using this migration tool. In this version, virtual extent information is created as special information. Applying the above formula, the start CCHH of virtual extent information is "00000000"
(Hexadecimal), so both CCs and HHs are 0, and as shown below, it is the same as the conversion formula from CCB to CCHHR when creating an unloaded version with this migration tool.

TT ÷n ---→ ｜ ｜ ｜ ---→ Quotient → Cylinder number (CC) Remainder → Track number (HH) R → As is → Record number (R) Next, as mentioned above, from the FBA DASD to the CKD
The above-mentioned items required to set data on DASD
The configuration of an embodiment of the present invention, which is a migration tool for creating an unloaded version of MT of the JSECOPY program, will be described based on FIG. 1 and with reference to other figures.

In FIG. 1, 1 is the FBA DASD, and blocks B-1, B-2, A-
1, C-1... exists, 2 is a data reading unit that reads the directory part and member part from DASD 1, and 3 is a CCB discrimination unit that discriminates the CCB read from the directory area. and set this as the relative CI number CC in the dataset.
4 is a CI number register, and 5 is a block number register, in which the relative CI number CC in the data set and the intra-CI block number B output from the CCB discriminator 3 are respectively output.
6 is an arithmetic unit that performs the calculation of equation (1) above, 7 is a cylinder register that holds the number of tracks per cylinder, and 8 is a cylinder number register. 9 is a track number register that holds the quotient of the above equation (1); 1 is a track number register that holds the remainder of the above equation (1);
0 is the connection part and is the cylinder number register 8.
output CC, track number register 9 output HH
11 is a record number register which holds the data output from the block number register 5 and outputs it as record number R, Reference numeral 12 denotes a writing unit which uses the number of tracks per cylinder outputted from the cylinder register 7 obtained from the operating system 16 and the virtual extent information generated by this program 18 as special information, and writes the corresponding
The 13 directory sections read from the DASD are used as directories, and the CCHHR and 14 member bodies output from the connecting section 10 are placed on the magnetic tape 15.
13 is the memory of the directory holding unit where the directory read from DASD 1 is held; 14 is the memory and the DASD
Blocks B-1 and B- of members read from 1
2. Those in which A-1, C-1... are held, 15
16 is a magnetic tape that stores the unloaded version of the JSECOPY program, and 16 is a magnetic tape that stores the unloaded version of the JSECOPY program.
17 issues an operating system macro (DEVTYPE) to obtain the number of tracks per cylinder of a DASD; 17 issues an operating system macro (NOTE) to obtain the CCB of a member block; 1
Reference numeral 8 denotes a virtual extent information holding unit on memory generated by this migration tool.

The operation of the configuration of one embodiment of the present invention shown in FIG. 1 will be explained.

(1) Operating system macro issuing unit 1
In step 6, issue the DEVTYPE macro to obtain the number of tracks per cylinder of FBA DASD and set the number of tracks per cylinder.
This value is stored in the register 7, and this value and the virtual extent information 18 generated by this migration tool are written to the magnetic tape 15 as special information. The reading unit 2 sequentially reads directories from the DASD 1 in which the FBA partitioned data set is stored, and the writing unit 15 writes them onto the magnetic tape 15 in the order in which they are received. The reading section 2 reads out the blocks of the member section one by one from the segmented data set. At this time, the macro issuing unit 17 of the operating system issues the NOTE macro, obtains the CCB of the read block, and
It is sent to the CCB discriminator 3.

(2) CCB discriminator 3 is the relative CI number within the dataset
CC and intra-CI block number B are identified, the CI number CC is set in the CI number register 4, and the block number B is set in the block number register 5.

(3) The arithmetic unit 6 calculates the above (( 1) Calculate the expression. Then, the quotient obtained from the calculation result is set in the cylinder number register 8, and the remainder is set in the track number register 9.

(4) The connecting unit 10 uses the quotient set in the cylinder number register 8 as the CKD cylinder number CC, and the remainder set in the track number register 9 as the CKD track number, which is also transmitted from the block number register 5. "CCHHR" is created by concatenating the record numbers in the record number register 11.

(5) The writing section 12 adds "CCHHR" output from the connecting section 10 to the member blocks B-1, B-2, A-1, etc. transmitted from the memory 14, and writes them onto the magnetic tape 15. Fill in.

In this way, you can automatically obtain the unloaded version of the JSECOPY program for FBA→CKD, so from now on, as shown in Figure 6,
CKD by JSECOPY dataset utility
We can obtain a partitioned data set for

Therefore, it is no longer necessary to bring the source to each system and perform compilation, combined editing, etc. to create a load module, as was the case in the past.
Modules can be transferred directly via magnetic tape.

〔Effect of the invention〕

According to the present invention, a partitioned data set for FBA can be created.
Since the JSECOPY program for CKD can be converted to an unloaded version, a large amount of load modules created in each system can be migrated as is, increasing data processing efficiency.

[Brief explanation of drawings]

Figure 1 is a configuration diagram of one embodiment of the present invention, Figure 2 is
Explanation diagram of files on DASD in FBA format, Part 3
Figure 4 is an explanatory diagram of a data set, Figure 4 is a detailed diagram of a partitioned data set, Figure 5 is an explanatory diagram of the block position of a CKD format DASD, and Figure 6 is the use of a migration tool using the address conversion method of the present invention. Fig. 7a is an explanatory diagram of the data set division state and virtual extent information in CKD, b is an explanatory diagram of the magnetic tape of the unloaded version of the JSECOPY program, and Fig. 8 is an explanatory diagram of the address conversion state. It is. In the figure, 1 is the FBA DASD, 2 is the readout section, and 3
is the CCB discrimination section, 4 is the CI number register, 5 is the block number register, 6 is the calculation section, 7 is the cylinder
8 is a cylinder number register, 9 is a track number register, 10 is a connection section, 11 is a record number register, 12 is a write section, 13 is a directory holding section, 14 is a member body holding section, 16 is a
Operating system macro (DEVTYPE) issuer for obtaining the number of tracks per cylinder of FBA DASD, 17 is a member block.
A macro (NOTE) issuing unit of the operating system for obtaining a CCB, and 18 indicate a virtual extent information holding unit generated by this migration tool.

Claims (1)

[Claims] 1. A fixed-length DASD in which fixed-length format data is stored, a reading unit that reads the directory and member part of a partitioned data set from the DASD, and a unit that obtains the number of tracks per cylinder of the DASD from the operating system. , a virtual extent information generation unit consisting of absolute addresses representing the start and end of the partitioned data set on the variable-length format DASD, the number of tracks, and the number of extents, and a generation unit for generating partitioned data set on the fixed-length format DASD from the operating system. an arithmetic unit that calculates the quotient and remainder by dividing the CCB, which is the address of the member part of and the cylinder number of the variable length DASD, which is the quotient, and the record number of the variable length DASD, which is the remainder, which is the track number and block number (part B of the CCB) of the variable length DASD. and the partitioned data set read from the DASD, using the number of tracks per cylinder of the DASD obtained from the operating system and the virtual extent information as special information, and using the directory part of the partitioned data set read from the DASD as a directory. The variable length format DASD obtained from the above calculation result is stored in each block of the member section of
between a fixed-length data set and a variable-length data set, characterized in that it has a reading unit that writes a block to which a cylinder number of a fixed-length data set, a track number of a variable-length data storage format DASD, and a record number of a variable-length data storage format data set are added to an external data storage unit. Address translation method.