Re: DFHSM question concerning Duplex copies
Wow...sounds like Dual Copy, but for tape. -- All the best, Scott T. Harder On 5/28/09, O'Brien, David W. (NIH/CIT) [C] obrie...@mail.nih.gov wrote: Thanks Allan, I finally found the section you were quoting and after reading concede your point. I've learned something so the day is not a total loss. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 4:38 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies I have not been actively duplexing in a while, but in my past life, assuming no other problems with the tape, if one or the other tapes of the primary/duplex pair hit end-of-volume, this would cause a new pair to start. From the z/OS 1.9 DFSMShsm Users Guide SC35-0421-07 Topic 3.4.2.1.1 When the original tape reaches its percent-full capacity, DFSMShsm must flush all the data to both tapes before performing a FEOV. If natural EOV is reached on either tape, DFSMShsm marks the tapes as FULL and restarts processing of the current data set from its beginning on two new tapes. IOW, the shorter tape wins. HTH, -Original Message- From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On Behalf Of O'Brien, David W. (NIH/CIT) [C] Sent: Thursday, May 28, 2009 2:21 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies From the HSM Guide: The alternate tape must have the same tape geometry as the original. For example, if the original tape is a 3590-1 tape, the alternate must be also. The question is: Is the tape geometry of the 9840A and 9840D the same? Allan, is your Duplex parameter defined as follows: SETSYS DUPLEX(MIGRATION(Y ERRALT(MARKFULL))) That would explain your following comment: Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. In my environment any problem with the duplex tape results in the function (Migrate, Backup or Recycle) continuing and an HSM internal Tapecopy task for each unduplex'd tape. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 3:03 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies This is actually the standard way duplexing works. Data is written simultaneously to both the primary and duplex copy. Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. Reading between the lines, it sounds like your management want's to use (e.g.) 1 primary (long) and 2 duplex (short) to make up a primary/duplex pair. This cannot be done. A 1 for 1 relationship between primary and duplex volumes is mandatory. IIRC, the primary/duplex pair must be the same device type as well. As stated above, the shorter tape controls. HTH, snip Is there a way to fool HSM into using only part of a tape for ML2 and Backup so that a shorter tape could be used for duplexing? We currently use 9840A drives for both local and remote (Duplex) ML2 and Backup tapes. Management would like to transfer the local HSM workload to an SL8500 with 9840D tape drives. Both the 9840As and 9840Ds are gen'd as emulated 3490s. We currently use the following Tapeutilization parameters: SETSYS - TAPEUTILIZATION(UNITTYPE(L9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) SETSYS - TAPEUTILIZATION(UNITTYPE(R9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) /snip -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists
Re: DFHSM question concerning Duplex copies
This is actually the standard way duplexing works. Data is written simultaneously to both the primary and duplex copy. Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. Reading between the lines, it sounds like your management want's to use (e.g.) 1 primary (long) and 2 duplex (short) to make up a primary/duplex pair. This cannot be done. A 1 for 1 relationship between primary and duplex volumes is mandatory. IIRC, the primary/duplex pair must be the same device type as well. As stated above, the shorter tape controls. HTH, snip Is there a way to fool HSM into using only part of a tape for ML2 and Backup so that a shorter tape could be used for duplexing? We currently use 9840A drives for both local and remote (Duplex) ML2 and Backup tapes. Management would like to transfer the local HSM workload to an SL8500 with 9840D tape drives. Both the 9840As and 9840Ds are gen'd as emulated 3490s. We currently use the following Tapeutilization parameters: SETSYS - TAPEUTILIZATION(UNITTYPE(L9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) SETSYS - TAPEUTILIZATION(UNITTYPE(R9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) /snip -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: DFHSM question concerning Duplex copies
From the HSM Guide: The alternate tape must have the same tape geometry as the original. For example, if the original tape is a 3590-1 tape, the alternate must be also. The question is: Is the tape geometry of the 9840A and 9840D the same? Allan, is your Duplex parameter defined as follows: SETSYS DUPLEX(MIGRATION(Y ERRALT(MARKFULL))) That would explain your following comment: Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. In my environment any problem with the duplex tape results in the function (Migrate, Backup or Recycle) continuing and an HSM internal Tapecopy task for each unduplex'd tape. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 3:03 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies This is actually the standard way duplexing works. Data is written simultaneously to both the primary and duplex copy. Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. Reading between the lines, it sounds like your management want's to use (e.g.) 1 primary (long) and 2 duplex (short) to make up a primary/duplex pair. This cannot be done. A 1 for 1 relationship between primary and duplex volumes is mandatory. IIRC, the primary/duplex pair must be the same device type as well. As stated above, the shorter tape controls. HTH, snip Is there a way to fool HSM into using only part of a tape for ML2 and Backup so that a shorter tape could be used for duplexing? We currently use 9840A drives for both local and remote (Duplex) ML2 and Backup tapes. Management would like to transfer the local HSM workload to an SL8500 with 9840D tape drives. Both the 9840As and 9840Ds are gen'd as emulated 3490s. We currently use the following Tapeutilization parameters: SETSYS - TAPEUTILIZATION(UNITTYPE(L9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) SETSYS - TAPEUTILIZATION(UNITTYPE(R9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) /snip -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: DFHSM question concerning Duplex copies
I have not been actively duplexing in a while, but in my past life, assuming no other problems with the tape, if one or the other tapes of the primary/duplex pair hit end-of-volume, this would cause a new pair to start. From the z/OS 1.9 DFSMShsm Users Guide SC35-0421-07 Topic 3.4.2.1.1 When the original tape reaches its percent-full capacity, DFSMShsm must flush all the data to both tapes before performing a FEOV. If natural EOV is reached on either tape, DFSMShsm marks the tapes as FULL and restarts processing of the current data set from its beginning on two new tapes. IOW, the shorter tape wins. HTH, -Original Message- From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On Behalf Of O'Brien, David W. (NIH/CIT) [C] Sent: Thursday, May 28, 2009 2:21 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies From the HSM Guide: The alternate tape must have the same tape geometry as the original. For example, if the original tape is a 3590-1 tape, the alternate must be also. The question is: Is the tape geometry of the 9840A and 9840D the same? Allan, is your Duplex parameter defined as follows: SETSYS DUPLEX(MIGRATION(Y ERRALT(MARKFULL))) That would explain your following comment: Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. In my environment any problem with the duplex tape results in the function (Migrate, Backup or Recycle) continuing and an HSM internal Tapecopy task for each unduplex'd tape. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 3:03 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies This is actually the standard way duplexing works. Data is written simultaneously to both the primary and duplex copy. Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. Reading between the lines, it sounds like your management want's to use (e.g.) 1 primary (long) and 2 duplex (short) to make up a primary/duplex pair. This cannot be done. A 1 for 1 relationship between primary and duplex volumes is mandatory. IIRC, the primary/duplex pair must be the same device type as well. As stated above, the shorter tape controls. HTH, snip Is there a way to fool HSM into using only part of a tape for ML2 and Backup so that a shorter tape could be used for duplexing? We currently use 9840A drives for both local and remote (Duplex) ML2 and Backup tapes. Management would like to transfer the local HSM workload to an SL8500 with 9840D tape drives. Both the 9840As and 9840Ds are gen'd as emulated 3490s. We currently use the following Tapeutilization parameters: SETSYS - TAPEUTILIZATION(UNITTYPE(L9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) SETSYS - TAPEUTILIZATION(UNITTYPE(R9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) /snip -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
Re: DFHSM question concerning Duplex copies
Thanks Allan, I finally found the section you were quoting and after reading concede your point. I've learned something so the day is not a total loss. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 4:38 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies I have not been actively duplexing in a while, but in my past life, assuming no other problems with the tape, if one or the other tapes of the primary/duplex pair hit end-of-volume, this would cause a new pair to start. From the z/OS 1.9 DFSMShsm Users Guide SC35-0421-07 Topic 3.4.2.1.1 When the original tape reaches its percent-full capacity, DFSMShsm must flush all the data to both tapes before performing a FEOV. If natural EOV is reached on either tape, DFSMShsm marks the tapes as FULL and restarts processing of the current data set from its beginning on two new tapes. IOW, the shorter tape wins. HTH, -Original Message- From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On Behalf Of O'Brien, David W. (NIH/CIT) [C] Sent: Thursday, May 28, 2009 2:21 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies From the HSM Guide: The alternate tape must have the same tape geometry as the original. For example, if the original tape is a 3590-1 tape, the alternate must be also. The question is: Is the tape geometry of the 9840A and 9840D the same? Allan, is your Duplex parameter defined as follows: SETSYS DUPLEX(MIGRATION(Y ERRALT(MARKFULL))) That would explain your following comment: Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. In my environment any problem with the duplex tape results in the function (Migrate, Backup or Recycle) continuing and an HSM internal Tapecopy task for each unduplex'd tape. Dave O'Brien NIH Contractor From: IBM Mainframe Discussion List [ibm-m...@bama.ua.edu] On Behalf Of Staller, Allan [allan.stal...@kbm1.com] Sent: Thursday, May 28, 2009 3:03 PM To: IBM-MAIN@bama.ua.edu Subject: Re: DFHSM question concerning Duplex copies This is actually the standard way duplexing works. Data is written simultaneously to both the primary and duplex copy. Whichever tape is shorter will cause the initial primary/duplex pair to be marked full and a new pair created. Reading between the lines, it sounds like your management want's to use (e.g.) 1 primary (long) and 2 duplex (short) to make up a primary/duplex pair. This cannot be done. A 1 for 1 relationship between primary and duplex volumes is mandatory. IIRC, the primary/duplex pair must be the same device type as well. As stated above, the shorter tape controls. HTH, snip Is there a way to fool HSM into using only part of a tape for ML2 and Backup so that a shorter tape could be used for duplexing? We currently use 9840A drives for both local and remote (Duplex) ML2 and Backup tapes. Management would like to transfer the local HSM workload to an SL8500 with 9840D tape drives. Both the 9840As and 9840Ds are gen'd as emulated 3490s. We currently use the following Tapeutilization parameters: SETSYS - TAPEUTILIZATION(UNITTYPE(L9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) SETSYS - TAPEUTILIZATION(UNITTYPE(R9840) PERCENTFULL(2200)) - TAPESPANSIZE(1000) /snip -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
