DFO Life Span

[Douglass Anderson]

How long has DFO for VMS been on the market?

[Volker Halle · Accepted solution]

Douglass,

support for the OpenVMS XQP MOVEFILE primitive was introduced with OpenVMS Alpha V6.1 in 1994. The MOVEFILE source module started life in 1990. Defragmenters were a very hot topic back then and I seem to remember discussions at Digital, whether a Defragmenter should be released by DEC, before there was a secure and atomic move file primitive available in the XQP.

The oldest public reference I could find was to POLYCENTER File Optimizer V2.0 (Kit: DFG020) on some Campus CDs from 1994.

Volker.

[Kris Clippeleyr]

Hi,

Don't exactly know, but 2.9 is current, 2.7 dates from 2002, and 2.6 from 2001.
I heard about it for the first time in the 90s. I found a copyright notice on DFO 2.3, stating:
"© Digital Equipment Corporation 1991, 1997. All rights reserved."
So it's been around for quite a while.

Greetz,
Kris (aka Qkcl

[John Gillings]

Douglass,

You're asking a lot of questions about DFO... I'm just wondering why it's such a big deal. Maybe you should talk to a sales consultant about your exact requirements?

Some things to think about:

Disk fragmentation on OpenVMS Files-11 volumes is (generally) NOT a huge problem. The file system is designed to tolerate moderate levels of fragmentation without excessive performance penalty. There are also many measures available to avoid fragmentation.

The heart of defragmenting is the MOVEFILE XQP primitive. This is part of OpenVMS, and it's where most of the "risk" of defragging disks lies. All defrag products should use MOVEFILE to do the actual work. The rest is really just "interface". Choosing which files to move and, perhaps, trying to position files intelligently based on usage.

Before MOVEFILE, commercial defraggers were probably THE most common source of file and disk corruption. Since MOVEFILE, the risky part of the operation has been pushed back onto OpenVMS itself, and all defrag products that use it are equally safe.

As Volker has pointed out, the product has been around about a decade, but perhaps more importantly, it's service and support record is very good. It's a very low noise product - very low volume of service calls relative to its customer base. Most issues reported against it are questions of usage or performance. Bugs are very rare, and I'm not aware of any instances of data loss attributable to DFO.

In choosing a product, select the one which meets your needs, has the best interface from your perspective, and is within your budget. But, keep in mind that many people believe that a disk defragmentor is NOT necessarily a vital component of the OpenVMS System Manager's toolkit.

[comarow]

It depends on the application and if it changes.

File fragmentation results in window turns and extra I/O. It can ruin the results of autogen when it can't create reasonably contiguous files. It generally runs in the background.

Otherwise, backup and restore certainly do a perfect job of defragmenting disks.

Certainly, large cluster sizes, preallocating file space will help reduce fragmentation.

Here is the URL for the spd.

http://h18000.www1.hp.com/info/SP5585/SP5585PF.PDF

[Ian Miller.]

Douglass, I see you have not assigned any points yet. Have a look at
http://forums1.itrc.hp.com/service/forums/helptips.do?#33

and assign 0 to 10 points to each response.

[Veli Körkkö]

DFG V1.0 was on consolidated distribution on 1992 March timeframe. I believe MOVEFILE was on VMS V5.5-x. I specifically remember running it
on VAX/VMS V5.5-2 and having an issue where
SYS.EXE did not have MOVEFILE DISABLED, got moved around, we had $SET TIME being executed peridocially and got corrupted files.

So, I'd say DFG/DFO/PFO has been around well over 10 years. Initial versions used RDB (for the database containing so called scripts), later it changed to using RMS for the database.

_veli

[Douglass Anderson]

solved