SWAPDUMP, SIMSWAP, SIMSWAPMON Statements for Remote SWAPDUMP via TCP/IP

This section shows operands of the SWAPDUMP, SIMSWAP, and SIMSWAPMON statements that are only used for remote operations, or have special meanings or considerations on remote operations, or are particularly important on remote operations. Many of the standard operands in section SWAP, SWAPDUMP, SIMSWAP, SIMSWAPMON Statements can also be used.

SWAPDUMP, SIMSWAP, and SIMSWAPMON Statement Operands

TYPE=FULL

Must be specified on the SWAPDUMP, SIMSWAP, or SIMSWAPMON statement.

ALLOWPAV=

APV=

The ALLOWPAV= operand specifies whether the Parallel Access Volume (PAV) aliases for the source volume are disabled during the SWAPDUMP operation. (On a remote operation (REMOTEIP=), the PAV aliases for the target devices at the remote site are never disabled.)

NO

NO - Indicates that the Parallel Access Volume (PAV) aliases for the source volume are disabled at the start of the SWAPDUMP operation and re-enabled after the SWAPDUMP operation completes.

YES

Indicates that the Parallel Access Volume (PAV) aliases for the source volume stay enabled during the SWAPDUMP operation.

Default: NO.

ALLOWPAV=YES is strongly recommended for remote SWAPDUMP. Since the operation is likely to run for a long time (days), it is important to continue to use facilities that enhance performance.

ALLOWPAV=YES needs to be specified on all of the SWAPDUMP and MONITOR tasks that you wish to leave PAV active.

ALLOWZHPF=

NO

At the beginning of a SWAPDUMP, FDRPAS temporarily disables zHPF on the volume being processed; FDRPAS turns zHPF back on at the end of the SWAPDUMP.YES

FDRPAS does not disable zHPF during the SWAPDUMP process. Transport Control Word (TCW) commands used with zHPF are allowed throughout the operation.

Default: NO.

ALLOWZHPF=YES is strongly recommended for remote SWAPDUMP. Since the operation is likely to run for a long time (days), it is important to continue to use facilities that enhance performance.

ALLOWZHPF=YES needs to be specified on all of the SWAPDUMP and MONITOR tasks that you wish to leave zHPF active.

BUFNO=

nn

Specifies the number of I/O buffers that FDRPAS uses while copying data from the source volume to the target device, from 4 to 32. On local operations the default is BUFNO=32, but on remote operations (indicated by REMOTEIP=) the default is BUFNO=8. With this default, 4 tracks worth of data, or up to about 200K, will be sent to TCP/IP for each transmission. In most cases, this should give a good balance between the efficiency of larger transmissions and the risks of overloading the connection. The best value for a given environment depends on several factors, including how much real memory is available to back the buffers, how many simultaneous tasks are specified (MAXTASKS=), and the bandwidth of the TCP/IP connection.

MAX

Equivalent to 32.

If BUFNO=nn is specified, nn should be an even number and is rounded up if odd. FDRPAS divides this buffer set in half, in order to overlap input and output I/Os. If REMOTEIP= and BUFNO= are both specified, BUFNO= must be specified after REMOTEIP= to be honored.

Default: With REMOTEIP=, the default is 8. (Otherwise, the default is 32.)

CPYVOLID=YES

On remote operations, CPYVOLID=YES is forced, whether it is specified or not. It is expected that the objective is to create a usable system in which the target volumes can be brought online; therefore the volume labels of the target volumes are set to VOL1.

KEEPACTIVE=

REPEAT

After processing each volume, SWAPDUMP does not remove the I/O intercepts, but leaves them installed to continue to monitor updates. On remote operations, it is expected that KEEPACTIVE=REPEAT will be specified, so that all of the SWAPDUMPs can be completed in a synchronized manner. After processing all of the volumes specified by MOUNT commands, SWAPDUMP does not terminate, but waits for 30 seconds and then process all of the volumes again. The first cycle through all of the volumes is the initial copy cycle, and may take a long time if there are many volumes. Later cycles only recopy updated tracks, and are much shorter. The process continues until a MODIFY jobname,KEEPACTIVE=DONE console command is issued. This command may be issued manually when the installation is ready to complete the migration, or it may be issued automatically by a MONITOR TYPE=KEEPACTIVE job. When KEEPACTIVE=DONE is issued, SWAPDUMP terminates the current cycle after the copy of the currently active volumes, performs one final cycle, and completes the operation for all of the volumes.

NO

FDRPAS completes the operation for each source volume as soon as the source and target devices are in synchronization, without waiting for any confirmation.

KEEPACTIVE=NO can be used as a cleanup function if a KEEPACTIVE=REPEAT operation is terminated by a failure or by a STOP(P) or CANCEL(C) command, leaving the I/O intercepts installed, and for some reason it is not desired to resume the operation. With KEEPACTIVE=NO, FDRPAS performs one cycle through all of the volumes, completes the operation, and deinstalls the I/O intercepts.

When using KEEPACTIVE=NO as a cleanup function, any volumes that have not completed the initial copy phase, or have failed, go through the initial copy phase. This is not useful if all you want to do is clean up the I/O intercepts. In this situation, contact INNOVATION for an alternative cleanup function.

Default: NO.

LARGERSIZE=

NO

The output device must have the same number of data cylinders as the source volume. The output device does not have to have the same number of alternate cylinders as the source volume. If the number of alternates is different, FDRPAS corrects the alternate count after the swap.

OK

The target device may have more data cylinders than the source volume. LARGERSIZE=OK allows you to migrate from one model of a DASD volume to a larger model. For example, you can migrate from a 3390-3 (3339 cylinders) to a 3390-9 (10017 cylinders). The volume size is updated in the VTOC and VTOCIX (if active). Note that if the target device is larger than the source, you receive an FDRW66 message with the “OVERRIDE OF WARNINGS” text.

Since the target device remains offline when doing a SWAPDUMP, the VTOCIX (if present) is disabled and the “DOS” flag is set in the F4 DSCB if the target device is a larger DASD volume. If the target volume is subsequently varied online to some LPAR, then the first new data set allocation causes the free space to be updated, but the VTOCIX remains disabled. To refresh the VTOCIX you must use FDRPAS “SWAPBUILDIX” (see SWAPBUILDIX and EXPANDVTOC Statement) or ICKDSF “BUILDIX”.

Default: NO.

LARGESWAP=

nnnnn

Specifies a value slightly larger than the number of volumes being swapped, from 2000 to 16000. The purpose of this operand is to reduce the below-the-line CSA requirements. This operand should be used when FDRPAS is processing more than 2000 volumes at the same time with KEEPACTIVE=REPEAT, either in one job or in multiple jobs. For example, for 5000 volumes, specify LARGESWAP=6000. When LARGESWAP= specifies a value larger than the number of volumes being swapped, the requirement for below-the-line CSA is no more than 400 KB.

LARGESWAP= needs to be specified on all of the SWAP and MONITOR tasks running in the same time period.

MAXCARDS=

nnnn

Specifies the maximum number of MOUNT statements that can be present in this FDRPAS step, from 1 to 9999.

Default: 250.

MAXTASKS=

nnn

Specifies the maximum number of volumes that can be processed concurrently by this SWAPDUMP job; from 1 to 100 (from 32 to 100 for SIMSWAPMON). You must follow the SWAP statement with multiple MOUNT statements, specifying the volumes to be processed. If the number of MOUNT statements exceeds MAXTASKS=nnn, as is expected with KEEPACTIVE=REPEAT, FDRPAS starts the indicated number of SWAPDUMP tasks; as each one finishes another one is started, until all MOUNT statements have been processed. With KEEPACTIVE=REPEAT, FDRPAS then pauses briefly and starts over from the beginning of the list of volumes, processing the indicated number of volumes concurrently, recopying tracks that have been updated.

In order to separate the messages from these SWAPDUMPs, the messages are written to SYSPRIxx DD statements. These statements are dynamically allocated as SYSOUT=* if nor provided in the JCL. At the termination of each SWAPDUMP subtask, most of its messages are also written to SYSPRINT so that all FDRPAS messages are in one place.

Default: SIMSWAPMON defaults to its minimum MAXTASKS= value of 32, but a value up to 100 may be specified. For SIMSWAP and SWAPDUMP, the default is that only one volume is processed at a time. If multiple MOUNT statements are provided, they are processed serially, and all messages are written only to SYSPRINT.

For remote SWAPDUMP via TCP/IP, when a single SWAPDUMP job is copying all volumes at the installation, it is expected that MAXTASKS=64 will commonly be used.

For SIMSWAPMON, code the MAXTASKS= value that you intend on using for the actual SWAP or SWAPDUMP process to allow the SIMSWAPMON task to validate the value.

NOTIFYERR=

userid

Specifies a TSO user ID. Error messages from FDRPAS are sent to that user ID using the SEND command, for immediate notification of problems in FDRPAS execution. If the TSO user is logged on to the same LPAR as the FDRPAS SWAPDUMP task, the messages usually appear as soon as the user presses Enter. On another LPAR, the user must use the LISTBC command to view them, or they are displayed when the user next logs on. The messages also appear in the job log in SYSOUT.

Default: No TSO notification.

For related information, see Recovering From Errors in Remote SWAPDUMP.

REMOTEIP=

Specifies that this operation is a SWAPDUMP to a remote site using TCP/IP. The target devices are not attached to the system on which FDRPAS is running; FDRPAS sends the data from the source devices over communication lines to FDRPAS for TCP/IP at another location, which then writes the data to target devices at that location.

SUPERTARGET= must be specified to indicate a device that the SWAPDUMP job uses to communicate with the monitors.

nnn.nnn.nnn.nnn[..ppppp]

Specifies the TCP/IP address and port number that FDRPAS at the local site is to use when communicating with FDRPAS for TCP/IP at the remote site. The TCP/IP address consists of 4 numbers separated by periods; each number is from 0 to 255. The port number is one number from 0 to 65535, following the TCP/IP address and separated from it by two periods. If the port number is omitted, it defaults to 2017.

STOPAFTER=

Do not specify STOPAFTER= unless recommended.

mmmmm

Specifies a number of minutes, from 1 to 65535. This operand should only be used with KEEPACTIVE=REPEAT. If FDRPAS has not completed the initial copy pass of all of the volumes specified by MOUNT commands within this time period, then it will not begin new volumes within this cycle, but bypasses them with message FDR266. The volumes that are currently in progress will continue. When they complete, FDRPAS starts over from the beginning of the list of volumes. Since the earlier volumes have now completed the initial copy pass, they should complete much more quickly, allowing FDRPAS to reach the end of the list within the STOPAFTER period. It is also possible for STOPAFTER to take effect after the initial copy pass if update activity is very high.

The purpose is to limit the amount of memory needed for track update tables for earlier volumes.

Values higher than about 2800 are liable to cause failures, since there is a limit of 48 hours for how long a volume can go without a new pass being started.

If KEEPACTIVE=REPEAT and STOPAFTER= are both specified, STOPAFTER= must be specified after KEEPACTIVE=REPEAT to be honored.

Default: 1440 (24 hours).

SUPERTARGET=

ST=

This operand is required for remote operations using TCP/IP.

uuuu

Specifies the z/OS device address of the “supertarget” device. This must be specified as a 4-digit address, with a leading zero if required. This device is used as a work and communication area by the SWAPDUMP and monitor jobs; the same supertarget must be specified for the SWAPDUMP job and all of its corresponding monitors. The supertarget must be connected to the main SWAPDUMP LPAR and all of the monitor LPARs, must be offline on all of those LPARs, and should not contain any needed data. The supertarget does not need to be initialized in any way. One supertarget can be used by many SWAPDUMPs at the same time. The supertarget should be a 3390-9 or larger; it does not have to be as large as the volumes being processed.

SWAPID=

This operand is required (except under GENSWAP) when more than one SWAPDUMP or SIMSWAPMON job using just in time monitors is to run at the same time. (It is expected that remote SWAPDUMP operations are not under GENSWAP and always uses just in time monitors.) FDRPAS uses the SWAPID value to relate each SWAPDUMP or SIMSWAPMON job to its associated monitors. Without SWAPID, a SWAPDUMP or SIMSWAPMON job does not submit a just in time monitor to an LPAR where a monitor is already running. With SWAPID, the SWAPDUMP or SIMSWAPMON job submits the just in time monitor to an LPAR unless a monitor with a matching SWAPID is already running there.

nnn

Specifies a number from 1 to 255. Do not use 64; if 64 is specified, it is treated as 65. Each SWAPDUMP or SIMSWAPMON job that is to run at the same time must specify a different value.

Default: None. SWAPID is not needed when only one SWAPDUMP or SIMSWAPMON job runs at a time.

SWAPIOERR=

RETRY

During a remote SWAPDUMP operation, if an error occurs in TCP/IP communications, FDRPAS attempts to recover.

NORETRY

During a remote SWAPDUMP operation, if an error occurs in TCP/IP communications, FDRPAS immediately fails the SWAPDUMP of the volume being processed.

Default: NORETRY.

For related information, see Recovering From Errors in Remote SWAPDUMP.

SWAPIOERR=RETRY is strongly recommended for remote SWAPDUMP.

TRKTAB=

Do not specify TRKTAB=OLD unless recommended by INNOVATION.

The track table is an area in ECSA (common memory above the 16 MB line) where FDRPAS keeps a record of tracks that have been updated since the beginning of the run or since the end of the previous copy or update pass. The track table for a given volume is fully or partially obtained the first time the volume is updated after the beginning of a run or after the end of the previous copy or update pass, and is freed at the end of each copy or update pass.

NEW

The track table is a variable size that depends on update activity. Only 4 KB is obtained at a time.

OLD

The track table is a fixed size of 1 KB plus 2 times the number of data cylinders on the volume. When the volume is updated, the full size of the track table is obtained. This may be very large, especially for EAVs.

If KEEPACTIVE=REPEAT and TRKTAB= are both specified, TRKTAB= must be specified after KEEPACTIVE=REPEAT to be honored.

Default: With KEEPACTIVE=REPEAT, the default is TRKTAB=NEW; otherwise the default is TRKTAB=OLD.