It is known that compressing "uncompressable" data increases the size of the
object.
VMARC compresses, but CMSDDR already did
(you write: "41 tracks not compacted", meaning you used OUTPUT xxx
(COMPACT).
If you still prefer VMARC as opposed to COPYFILE PACK, use VMARC PACK xxx
(STORE, telling VMARC not to compress.
2009/10/16 Henry, Bob <[email protected]>
> I’ll look into that – thanks.
>
>
>
> *From:* The IBM z/VM Operating System [mailto:[email protected]] *On
> Behalf Of *Scott Rohling
> *Sent:* Thursday, October 15, 2009 7:36 PM
> *To:* [email protected]
> *Subject:* Re: CMSDDR & VMARC question
>
>
>
> (actually - maybe the PIPEDDR DUMP/RESTORE over TCPIP is exactly what you
> want? if the zVM systems have connectivity, this lets you dump a disk
> directly from one system to another - and will save you steps...)
>
> Scott
>
> On Thu, Oct 15, 2009 at 5:34 PM, Scott Rohling <[email protected]>
> wrote:
>
> Maybe try a COPYFILE (PACK of the output file instead of using VMARC?
> That way if you use F 1024 when FTPing to zVM, you can just COPYFILE (UNPACK
> it.. I'm just assuming you were using VMARC to help ensure the xfer was
> blocked correctly.. Using COPYFILE (PACK accomplishes the same thing.
> Just not sure how they compare on 'compression'...
>
> You might also look at PIPEDDR -- and the PACK option .. (available on the
> zVM download page. google 'zvm download' to find it).
>
> Scott
>
>
>
> On Thu, Oct 15, 2009 at 5:18 PM, Henry, Bob <[email protected]>
> wrote:
>
> I was going to do that, but some of the MDs are VSE full volume disks.
>
>
>
> *From:* The IBM z/VM Operating System [mailto:[email protected]] *On
> Behalf Of *Scott Rohling
> *Sent:* Thursday, October 15, 2009 7:17 PM
> *To:* [email protected]
> *Subject:* Re: CMSDDR & VMARC question
>
>
>
> Don't know about VMARC -- but CMSDDR is packing the entire minidisk - empty
> space and all... Why not just VMARC the files on the disk and forget
> about CMSDDR?
>
> Scott
>
> On Thu, Oct 15, 2009 at 5:05 PM, Henry, Bob <[email protected]>
> wrote:
>
> I’m using CMSDDR and VMARC to transfer some CMS minidisks via FTP. Both
> utilities seem to produce more output than the original data on the
> minidisk(s). Here’s an example.
>
>
>
> User has 25 files (mostly COBOL source and/or JCL) in a 4 cylinder 3390
> minidisk, blocked 4096. A “Q DISK” shows 76 blocks used, 644 blocks left.
>
> CMSDDR on that minidisk shows 2,957,040 bytes IN, 2,634,392 bytes OUT, 41
> tracks not compacted. The output file from CMSDDR has 101 records of
> LRECL=49152 using 644 blocks (size 4096).
>
> VMARC (of the CMSDDR output file) shows IN=2,634,392 and OUT=3,142,240. It
> produces a file of 38,736 records with LRECL=80 using 757 blocks (size
> 4096).
>
>
>
> Does anyone have any explanation why these utilities “grow” the amount of
> data to be transmitted rather than “shrinking” it? Am I missing something?
> I’m using just the default options for both programs.
>
>
>
> Any help would be appreciated.
>
>
>
>
>
>
>
--
Kris Buelens,
IBM Belgium, VM customer support
