|
View previous topic :: View next topic
|
| Author |
Message |
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
Hello everyone!
I have a question about HSM duplex tapes and the recycle process, so if anyone knows the answer please respond.
The goal is to automaticaly change the storage group of recycled duplex tapes (to scratch), so they can be reused as duplex tapes for new ML2 tapes.
Since DFSMSrmm isn't available, I was thinking a scenario along these lines would be the answer:
1. Operator issues the RECYCLE command (for ML2 tapes)
2. Upon marking a ML2 tape as empty and discarding it's duplex pair, HSM passes control to ARCTVEXT exit
3. ARCTVEXT exit submits a JCL job (using program ACBQBAI9 - Storage group volume add/Delete) which changes the storage group of the duplex tape to scratch
4. Profit
Ideas? Help? |
|
| Back to top |
|
 |
expat
Global Moderator
Joined: 14 Mar 2007
Posts: 8796
Location: Welsh Wales
|
|
|
|
| If the storage environment is set up correctly the duplex tapes should be set back to scratch status when the primary volume is scratched. |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
| expat wrote: |
| If the storage environment is set up correctly the duplex tapes should be set back to scratch status when the primary volume is scratched. |
Hi expat!
Do you maybe have any hints on where to look/what to check?
I tried the following HSM configuration:
SETSYS TAPEDELETION( -
BACKUP(HSMTAPE) -
MIGRATION(SCRATCH) -
DUMP(HSMTAPE))
But, after the primary volume is scratched (marked empty in MCDS, but storage group remains HSMML2), the duplex tape also remains in the HSMML2 storage group.
Upon writting to a new empty ML2 tape (from HSMML2 SG) HSM takes a new tape from the *SCRATCH* storage group for duplex.. |
|
| Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007
Posts: 8796
Location: Welsh Wales
|
|
|
|
Eeeek. I can't remember the settings, but you do have ARCTVEXT (I think) active.
Which TMS are you using. |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
Nope, ARCTVEXT isn't active at the moment.
It needs to be activated with a SETSYS command:
Does it make sense to activate it if there is no (non-IBM) TMS to use it?
| expat wrote: |
| Which TMS are you using. |
No TMS is available (AFAIK), neither DFSMSrmm (as mentioned) nor CA 1.. |
|
| Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007
Posts: 8796
Location: Welsh Wales
|
|
|
|
| Eeeeek, I've only ever used HSM with a TMS, which then does the required to put the tape back into scratch status. |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
Any suggestions where else to ask? 
(thanks for the responses so far!) |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
Hello,
Suggest you talk with the storage management people and learn how they keep track of the tape inventory.
I do not recall the last time i worked somewhere that had NO tape management system . . . And that covers many years and many installations. |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
| dick scherrer wrote: |
| Suggest you talk with the storage management people and learn how they keep track of the tape inventory. |
Excel spreadsheets
DFSMSrmm and CA 1 are not used, I am sure of that. |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
Hello,
Then i suspect the tape inventory is Very small or it is quite out of sync with the actual usage. . .
How do the volumes get recorded for every tape created by the system (in a timely manner?
| Quote: |
| DFSMSrmm and CA 1 are not used, I am sure of that. |
Is this something you "know" or did you ask the storage management people? There are several tape management systems from different vendors. . . |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
Hello dick,
| dick scherrer wrote: |
| Then i suspect the tape inventory is Very small or it is quite out of sync with the actual usage. . . |
considering the usage of the tape inventory, there isn't much to track.
DFSMShsm is the biggest consumer of it (with backup, dump and ML2 tapes), and there are some job-driven usage of the WORM tapes.
| dick scherrer wrote: |
| How do the volumes get recorded for every tape created by the system (in a timely manner? |
I'm not following on this one, could you please explain what you've ment? 
If you mean how it is known which tape is used for which purpose - most tapes are pre-defined and used in certain (name) ranges.
| dick scherrer wrote: |
| Is this something you "know" or did you ask the storage management people? |
This is something I "know" and also have asked the S.M. people. |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
Hello,
| Quote: |
| Quote: |
| How do the volumes get recorded for every tape created by the system (in a timely manner? |
I'm not following on this one, could you please explain what you've ment?
If you mean how it is known which tape is used for which purpose - most tapes are pre-defined and used in certain (name) ranges. |
Usually, scratch tapes are allocated dynamically from a "scratch pool". The volser(s) are not known before the job runs and creates the tape output.
If you jobs already know the volser(s), there may be issues when a potentially 5 volume tape only goes to 3 volumes. The other 2 volumes are kinda "lost" to the system.
If your system uses very little tape and it works this way, i'd probably leave it alone until a complete replacement of some kind is done. |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
| dick scherrer wrote: |
| Usually, scratch tapes are allocated dynamically from a "scratch pool". The volser(s) are not known before the job runs and creates the tape output. |
We have drifted away from the subject a bit
But, nevertheless - we're using "scratch pool" tapes only for ML2 DUPLEX tapes (since HSM is only able to take new duplex tapes from that pool). All other tapes are pre-defined.
It all works fine, with the exception of the mentioned problem, because a manual intervention (putting back unused duplex tapes to the "scratch pool") is needed every couple of days (and no, complete replacement isn't coming  ) |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
| Quote: |
| We have drifted away from the subject a bit |
Yup, probably my bad.
I reacted because one of my clients did a kinda "manual" tape library and i was tasked to convert the tape inventory to virtual tape instead of carts.
The research showed that the inventory was out of sync by 30+%. And the inventory was rather small - less than 10,000 volumes.
Sorry 'bout the distraction . . . 
d |
|
| Back to top |
|
|
Ivan P
New User
Joined: 08 Jun 2009
Posts: 63
Location: Europe
|
|
|
|
We also have a rather small inventory then.. (even less than 5,000 volumes).
Suggestions/Ideas on the solution or where else to ask are still welcome
Also help with the job (from step 3) which uses program ACBQBAI9 is wanted.
I can't seem to make it change the storage group.
Here's the used JCL:
| Code: |
//MYJOB JOB CARD ...,TIME=(0,10)
//MYLIB JCLLIB ORDER=SYS1.SACBCNTL
//ADDVOL1 EXEC ACBJBAOB,
// PLIB1='SYS1.DGTPLIB',
// TABL2=MYUSER.TEST.ISPTABL
//SYSUDUMP DD SYSOUT=*
//TEMPFILE DD DSN=&&VOLADDS,DISP=(NEW,PASS),
// SPACE=(TRK,(1,1)),LRECL=300,RECFM=F,BLKSIZE=300
//SYSTSIN DD *
ISPSTART CMD(ACBQBAI9) +
BATSCRW(132) BATSCRD(27) BREDIMAX(3) BDISPMAX(99999999)
/*
//VOLADD DD *
/*
//VOLALT DD *
UPDHLVLSCDS()
SCDSNAME(HLQ.MY.SCDS) VOL(MYVOL) SG(*SCRTCH*) STATUS(ENABLE)
/*
//VOLDEL DD *
/* |
SYSTSPRT list says:
| Code: |
READY
ISPSTART CMD(ACBQBAI9) BATSCRW(132) BATSCRD(27) BREDIMAX(3) BDISPMAX(99999999)
No Volume Delete records Specified
No Volume Add records Specified
MYUSER.MYJOB.JOB37134.D0000108.? was preallocated (no free was done).
READY
END |
And ISPLOG list says:
| Code: |
Time *** ISPF transaction log *** Userid: MYUSER Date: 12/06/29 Page: 1
10:13 Start of ISPF Log - - - - Session # 1 -------------------------------------------------------
10:13 TSO - Command - - ACBQBAI9
10:13 End of ISPF Log - - - - - Session # 1 ------------------------------------------------------- |
And still... "MYVOL" remains in the original storage group... |
|
| Back to top |
|
|
|
|
