30.09.2012 Views

Hot Topics - Messmer The Brain House

Hot Topics - Messmer The Brain House

Hot Topics - Messmer The Brain House

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Storage made easy<br />

Answers that won’t drive you nuts<br />

BY EDWARD BAKER AND SEAN MCMILLEN<br />

z/OS system programmers, z/OS storage<br />

administrators and independent software<br />

vendors are asking IBM some typical<br />

questions about storage. Let’s take a look:<br />

Dear IBM,<br />

I use tape mount management (TMM)<br />

heavily in my shop. This is in order to<br />

minimize the number of tape mounts and<br />

maximize the usage of my tape resources<br />

during tape mounts. I have noticed<br />

that these tape mount management<br />

data sets are not eligible for the fast<br />

subsequent migration (FSM) function. Is<br />

there anything that one can do to allow<br />

these data sets to be candidates for fast<br />

subsequent migration?<br />

Thanks,<br />

Mr. Tapemount<br />

Dear Mr. Tapemount,<br />

You are in luck; beginning in z/OS<br />

R7 we have implemented changes to fast<br />

subsequent migration that can solve your<br />

common dilemma. As you’re aware, fast<br />

subsequent migration is a function that<br />

allows data sets that have been recalled<br />

from migration level 2 (ML2) tape to be<br />

reconnected to the initial place of origin<br />

on that ML2 tape. This is only possible if<br />

the data set was not modified since being<br />

recalled. In a tape mount management<br />

(TMM) environment, data sets are<br />

normally assigned to a management class<br />

or storage group that does not require<br />

backup. You were most likely encountering<br />

this problem because no valid backup copy<br />

typically exists for data sets in a TMM<br />

environment.<br />

In z/OS V1R7, DFSMShsm, with<br />

assistance from other z/OS DFSMS<br />

components, developed a more rigorous<br />

FSM design that allows TMM-type data<br />

to be reconnected. In fact, even data<br />

sets without a valid backup copy are now<br />

eligible for fast subsequent migration as<br />

long as the data set remains unaltered<br />

since the recall. Let us know how this<br />

enhancement works for you!<br />

Thanks,<br />

IBMer<br />

To DFSMShsm Development,<br />

I utilize your HMIGRATE command<br />

daily. Typically, it’s when I am about to<br />

leave work and I want to migrate all my<br />

L0 volume data sets that I used that day.<br />

I migrate most of my data sets to ML1<br />

DASD, or even sometimes ML2 tape.<br />

When I found out that I could use partially<br />

qualified data set names on the command<br />

with the use of filters (also known as<br />

wildcards) I was ecstatic. This allowed me<br />

to accomplish all migrations with a single<br />

command. Because I don’t reuse every<br />

single data set created every day, I often<br />

find that I get an annoying error message.<br />

<strong>The</strong> error message appears for every data<br />

set that meets the filter criteria that is<br />

already migrated. Is there a way to<br />

turn off notification of all these error<br />

messages? I really like using the<br />

HMIGRATE with wildcards (*).<br />

Thank you,<br />

Mrs. Migrate<br />

Dear Mrs. Migrate,<br />

We understand your request, and<br />

we have an answer for it beginning<br />

in DFSMShsm V1R7. We will no<br />

longer issue the ARC1224I message for<br />

data sets already migrated to ML1 when<br />

you use the HMIGRATE command with<br />

partially qualified data set names and<br />

filters.<br />

Additionally, you will no longer see<br />

the ARC1282I error message when the<br />

data set already resides on an ML2 tape<br />

volume. This eliminates the multitude of<br />

error messages that you receive during<br />

your migration processing. By the way, we<br />

also implemented the bypassing of error<br />

messages whether or not you specify the<br />

ML2 parameter on your HMIGRATE<br />

command. Try it out and let us know what<br />

you think.<br />

Thanks,<br />

HSM Development<br />

To IBM,<br />

We have recently invested in your<br />

latest high capacity tape media, the IBM<br />

TotalStorage® 3592 Tape Cartridges. I<br />

find it just phenomenal that you can put<br />

300 GB of uncompressed data on a single<br />

tape these days. I remember back when the<br />

3480s came out in 1984 and the capacity<br />

was 200 MB of data. Back in those days<br />

that was a whole lot of data (ha-ha!).<br />

Anyway, we have found that DFSMShsm<br />

has not been filling our larger tapes to<br />

their full capacity. We see this primarily<br />

when we place hundreds of thousands<br />

February 2006 z/OS HOT TOPICS Newsletter, Issue 14 49

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!