Problems with CD/DVD drives on OpenVMS 8.3

[Christopher Blackburn]

Has anyone else had problems accessing files held on a CD/DVD from OpenVMS ?
My particular problem is a DVD produced on a PC as a ISO 9660 complient volume which I am trying to copy files from onto an OpenVMS server (RX2660). The DVD is a LiteOn connected to the RX2660 as an external USB device.
Once the copy operation is started it takes a very long time to actually copy the file. The only process activity on the system is the occasional DNA3CACP process which is actually doing the copy operation from the device. If I use Control-T on the copy process it only increases the I/O count by 1 each time Control-T is used.
The I/O count for the actual device increases by hundreds of thousands while the copy operation is being performed (it does eventually complete but can take 30 minutes). The file being copied is approximately 4Mb is size. I have noticed the same problem when using CDs on VMS on both the IA64 and ALPHA platforms.

[Heuser-Hofmann]

Why don't you use a USB-stick and a freeware-tool (mtool) to transfer files to VMS?

Eberhard

[Duncan Morris]

Christopher,

you would do well to study Hoff's pages about using CD/DVD with openVMS.

http://64.223.189.234/node/12

and

http://64.223.189.234/node/28

In particular, note the information about 512-bte blocks versus 2048-byte blocks.

Additionally try searching this forum for DVD, where you will find quite a number of relevant threads.

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1151273


Duncan

[Heuser-Hofmann]

1. There is no 2048/512 blocking problem because the DNdriver solves this internally.
2. Did you apply this patch?
VMS83I_CDDVD-V0200.ZIPEXE
This installs a new and better dndriver.
3. Upgrade to 8.3-1H1 if possible.
Then you'll see USB2-speed instead of poor USB1.1 performance.

eberhard

[Christopher Blackburn]

Duncan,
I've read the notes from HoffmanLabs and the other FAQs and there is no mention of very slow performance when using CD/DVD media. I note the discussion of 512 byte and 2048 byte buffers and I believe the DVD device I am trying to use is 2048. This does not seem to be overridable at mount time. The files are actually copied from the device but only after many minutes/hours. It has just taken 30 minutes to copy a single 4MB file and I am now copying 8000 images which are all small (<50Kbytes). This appears to be copying 2 files every second.

Heuser,
This would be the ideal solution as I have a 4GB USB stick. I can't find any tools that will copy the files from the DVD onto the USB stick in Files11 format so they can be read from OpenVMS. Do you know of any ?

[Christopher Blackburn]

Heuser,
I have not applied that patch. I will find it and apply it to the system.

Many thanks - Chris.

[Heuser-Hofmann · Accepted solution]

USB-Sticks usually are FAT32 formatted and mtools is a program to read/copy/remove files from a FAT32 device under OVMS.

You'll find it here:

http://antinode.info/ftp/mtools/

Hope this helps
Eberhard

[Willem Grooters]

>> My particular problem is a DVD produced on a PC as a ISO 9660 complient volume which I am trying to copy files from onto an OpenVMS server (RX2660).

That should be no problem (I've done that quite often on an Alpha).

>> The DVD is a LiteOn connected to the RX2660 as an external USB device.

Well, THAT could well make a difference. AFAIK, the transfer-rate of USB (even 2.0) is low compared to the nomally used interfaces on CD's (SCSI or ATAPI). And there is the driver itself that could add some latency as well.

>> I have noticed the same problem when using CDs on VMS on both the IA64 and ALPHA platforms.

It could depend on the CD/DVD on which the image was burned. I have CD's where COPY of a large file is fast, and ones where copy of a small file takes considerable time. The major difference: burned on another system (and different software) ...

[Christopher Blackburn]

Eberhard,

Many thanks - I will download mtools and try to make a copy of DVD onto a USB drive and access the files from there.

Willem,

Thanks for comments. I believe you are correct. Some writing software seems to make a better job than others. I have no idea what was used to write this DVD (it was made in the USA by a third party company). It seems to work at a reasonable rate on a PC.

[Hoff]

It appears you have a bad drive or bad disk or bad media. "Bad" here including "incompatible".

You can try a similar four megabit (Mb) or (as I'd assume intended, four MB (megabyte)) copy from a different hunk of media as a test.

Speeds can and do vary by bus and by device and by media.

Yes, V8.3-1H1 and USB 2.0 might help here.

But there's an issue here beyond the device speed.

It would not surprise me to encounter a bad or slow drive here. Do not assume any compatibility among media, nor among any two arbitrary CD nor DVD devices. These devices are flaky, buggy and inconsistent. This includes via USB and via IDE. Media too can be mis-rated. Put another way, if things get weird, swap the drive model and swap the media source.

In one case, three quarters of the USB bus traffic seen with a USB DVD device from a top-tier vendor was error recovery traffic. The device was tossing staggering numbers of errors, and the Windows recording package was fielding and retrying -- it was surprising that no messages were logged to the user that might indicate the device was, well, junk; all that was obvious was just slow I/O and bus contention. Put another way, if things get weird, swap the drive. Or swap the media source.

Do not assume that a PC recording package records correct volume structures, too. Versions of Nero, for instance, have caused me various problems. ISO-9660 must be strict subset; that's probably the case here, as you're able to access the disk. Put another way, swap the recording package.

As for drive performance itself, the recording process can take 30 to 45 minutes for a DVD, depending on the device and the IDE/DQ or USB/DN bus and the recording details.

Here, I'd try a different USB DVD drive. Or do use the USB key disk, yes. Or load the file on a local Mac OS X or Linux or Windows box (qv: different drive) and lob the file over to the OpenVMS I64 box via ftp.

[Christopher Blackburn]

Hoff,
Many thanks for your reply (you are the God of VMS), You guessed right it was 4MB rather than Mb.
I too believe the media is not good as I have tried the same DVD in two different drives (one USB on the IA64 Blade and one internal IDE drive on an old Alpha DS10). I get the same poor performance on both systems with the same slow behaviour. Again no errors are logged on the device but the IO count for the device is very large after a simple file copy. I'm copying files from the DVD onto the OpenVMS server. I'm currently up to 10 million operations and I've only copied a few thousand small files.
I will close the thead now and try to re-create the DVD media on a standalone PC to improve the compatibility with the VMS drives.

Again - Many thanks for your help.

[Christopher Blackburn]

Problem believed to be bad media or incompatibility between creating system/software (PC probably running NERO) and the destination system (HP BLADE).

Thanks for all replies.

[Hoff]

I've found the CDburnerPro XP to be functional and reliable (and free), if you're operating on Microsoft Windows. Ensure you use ISO-9660:1988 strict subset.

If operating on OpenVMS, there are commercial DVD recording packages available for various versions. One of the folks replying here offers one.

OpenVMS V8.3 and later includes integrated DVD recording with the COPY /RECORDABLE_MEDIA command. V7.3-1 and later includes CD recording. This assumes you have an appropriate burner on your OpenVMS box, and there are some good options for IDE buses. For USB burners, get ECOs and/or V8.3-1H1.

And media is always interesting when dealing with CD and DVD recording; cheaper recordable media is worth what you pay for it, or even less.

Various of the recording tools have verification centrally because of the problems seen with various recordable media and with flaky drives or flaky drive firmware.