Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

OpenVMS and Mac NFS issues

247 views
Skip to first unread message

rcyoung

unread,
Mar 13, 2011, 3:36:46 PM3/13/11
to
Has anyone had experience with getting NFS file sharing to/from
OpenVMS (8.2) from/to the Mac (10.6.6)? I can get the Mac volume
mounted over on OpenVMS, but any attempt to do a "dir" or "copy"
results in a timeout. Here is the terminal output

On OpenVMS, the Mac volume is mounted

TCPIP> show proxy

VMS User_name Type User_ID Group_ID Host_name

YOUNG OND 501 501 *


TCPIP> mount DNFS0: /HOST="rcyoung" /path="/Users/rcyoung" /
automount=inactivity
=7-00:00:00.0/share/force
%TCPIP$DNFSMOUNT-S-MOUNTED, /Users/rcyoung mounted on _DNFS16:[000000]
TCPIP>

When you go to the OS, you can see the disk (aka DNFS16), but no "dir"
is possible (you timeout)

$ sh dev d

Device Device Error Volume Free Trans Mnt
Name Status Count Label Blocks Count Cnt
A2100N$DKA0: Online 0
A2100N$DKA100: Mounted 0 ALPHA820 844200 349 1
A2100N$DKA200: Mounted 0 AXPDATA2 796608 1 1
A2100N$DKA300: Mounted 0 AXPDATA3 997408 1 1
A2100N$DKA400: Online wrtlck 0
A2100N$DVA0: Online 0
DNFS0: Online 0
DNFS16: Mounted 0 rcyou$/Users 0 1 1
$ dir dnfs16:[000000]
%DIRECT-E-OPENIN, error opening DNFS16:[000000].;* as input
-RMS-E-DNF, directory not found
-SYSTEM-F-TIMEOUT, device timeout

JF Mezei

unread,
Mar 13, 2011, 4:49:05 PM3/13/11
to
rcyoung wrote:
> Has anyone had experience with getting NFS file sharing to/from
> OpenVMS (8.2) from/to the Mac (10.6.6)?

File protection is generally the big problem. And not obvious how to fix
properly. Not it doesn't generate the timeout errors you see.

> $ dir dnfs16:[000000]
> %DIRECT-E-OPENIN, error opening DNFS16:[000000].;* as input
> -RMS-E-DNF, directory not found
> -SYSTEM-F-TIMEOUT, device timeout


I can reproduce this if VMS creates more than 315 versions of temp.png

(perhaps it is a limit on number of files in a directory).

This problem surfaced with an upgrade of OS-X, it used to not be a
problem before. My VMS config has not changed.

I am not sure if you are hit with the same issue.

VAXman-

unread,
Mar 13, 2011, 6:51:56 PM3/13/11
to

Hmm. OK. First look at the Mac.

What's in your /etc/exports?

Try just:

/Users -alldirs -network xx.yy.zz.0 -mask 255.255.255.0

Use your network in lieu of xx.yy.zz.0

Make certain nfsd is running:

% sudo nfsd status
nfsd service is enabled
nfsd is running (pid 376, 8 threads)

Make sure that the nfsd sees the /etc/exports

% sudo nfsd restart checkexports verbose

Lastly, verify your GID/UID with 'ls -ln'

On my MacBook Pro, these are 20 and 501, respectively.

On VMS:

TCPIP> SET HOST mbp /ADDRESS=xx.yy.zz.nn

TCPIP> ADD PROXY <localVMSusername> /GID=20 /HOST=mbp /UID=501

Check that you can actually see your NFS server on your Mac and that you
can get to it. You may need to access and change things in your Mac's
System Preferences->Sharing.

$ TELNET/PORT=2049 mbp
%TELNET-I-TRYING, Trying ... xx.yy.zz.nn
%TELNET-I-SESSION, Session 01, host mbp, port 2049
helpNET-I-ESCAPE, Escape character is ^]
%TELNET-S-REMCLOSED, Remote connection closed
-TELNET-I-SESSION, Session 01, host mbp, port 2049
$

Then and only then, you can use your MOUNT command

TCPIP> MOUNT DNFS0: /HOST=mbp /PATH="/Users/VAXman" /STRUCTURE=5
%TCPIP$DNFSMOUNT-S-MOUNTED, /Users/VAXman mounted on _DNFS13:[000000]

$ DIRECTORY/SECURITY/WIDTH=(FILENAME:25,OWNER=12) DNFS13:[000000]

Directory DNFS13:[000000]

.bash_history;1 [DEFAULT] (RWD,RWD,D,D)
.CFUserTextEncoding;1 [DEFAULT] (RWD,RWD,D,D)
.cshrc;1 [DEFAULT] (RWD,RWD,RD,RD)
^.cups.DIR;1 insufficient privilege or object protection violation
.DS_Store;1 [DEFAULT] (RWD,RWD,RD,RD)
^.editix.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
.emacs;1 [DEFAULT] (RWED,RWED,RD,RD)
^.fontconfig.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
.lesshst;1 [DEFAULT] (RWD,RWD,D,D)
.profile;1 [DEFAULT] (RWD,RWD,RD,RD)
^.serverauth.1561;1 [DEFAULT] (RWD,RWD,D,D)
^.serverauth.2345;1 [DEFAULT] (RWD,RWD,D,D)
^.serverauth.3064;1 [DEFAULT] (RWD,RWD,D,D)
^.serverauth.644;1 [DEFAULT] (RWD,RWD,D,D)
^.serverauth.668;1 [DEFAULT] (RWD,RWD,D,D)
^.ssh.DIR;1 insufficient privilege or object protection violation
^.sysdb20.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
^.Trash.DIR;1 insufficient privilege or object protection violation
.viminfo;1 [DEFAULT] (RWD,RWD,D,D)
.Xauthority;1 [DEFAULT] (RWD,RWD,D,D)
.Xmodmap;1 [DEFAULT] (RWD,RWD,RD,RD)
192^.168^.2.2;1 [DEFAULT] (RWED,RWED,RD,RD)
auralmoon-streamer.;1 [DEFAULT] (RWED,RWED,RD,D)
auralmoon.;1 [DEFAULT] (RWED,RWED,RD,RD)
Bootcamp^_2010.DIR;1 insufficient privilege or object protection violation
c.mst;1 [DEFAULT] (RWD,RWD,RD,RD)
Desktop.DIR;1 insufficient privilege or object protection violation
Documents.DIR;1 insufficient privilege or object protection violation
Downloads.DIR;1 insufficient privilege or object protection violation
ds10l.;1 [DEFAULT] (RWED,RWED,RD,RD)
GLUCOFACTS^_Deluxe.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
GlucofactsLog.log;1 [DEFAULT] (RWD,RWD,RD,RD)
Library.DIR;1 insufficient privilege or object protection violation
Movies.DIR;1 insufficient privilege or object protection violation
Music.DIR;1 insufficient privilege or object protection violation
Pictures.DIR;1 insufficient privilege or object protection violation
Projects.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
Public.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
rx2600.;1 [DEFAULT] (RWED,RWED,RD,RD)
Sites.DIR;1 [DEFAULT] (RWED,RWED,RED,RED)
switchJDK.sh;1 [DEFAULT] (RWED,RWED,RD,RD)
tmesis.;1 [DEFAULT] (RWED,RWED,RD,RD)
UpdateStartup.log;1 [DEFAULT] (RWD,RWD,RD,RD)
xx.;1 [DEFAULT] (RWED,RWED,RD,RD)

Total of 44 files.

--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG

All your spirit rack abuses, come to haunt you back by day.
All your Byzantine excuses, given time, given you away.

rcyoung

unread,
Mar 13, 2011, 10:40:26 PM3/13/11
to
What is

/Users -alldirs -network xx.yy.zz.0 -mask 255.255.255.0

or more precisely, the "/Users" part? Have not seen that before.

JF Mezei

unread,
Mar 13, 2011, 10:48:45 PM3/13/11
to

on your mac:


cat /etc/exports

And you will see the config for the "disks" you are exporting.

rcyoung

unread,
Mar 13, 2011, 10:51:09 PM3/13/11
to
I am (at least a partial) idiot. I wondered what "/Users was"...I
should have reread the posting to see he was talking about the
"contents" of the /etc/export file. Just want to be sure I "catch
myself" on that one before everyone thinks I'm a dolt.....it's been a
LONG DAY :-)

JF Mezei

unread,
Mar 13, 2011, 10:58:03 PM3/13/11
to
rcyoung wrote:
> I am (at least a partial) idiot.


You'll be glad to know your are not alone. There are plenty of people
like you and me :-)

rcyoung

unread,
Mar 13, 2011, 11:04:21 PM3/13/11
to
"Me thinks" it was the /etc/exports file. Things are looking
better.....let me play a bit more with it and if all is golden, I'll
give you a thumbs up!

rcyoung

unread,
Mar 13, 2011, 11:12:00 PM3/13/11
to
A quick "related" question- I have run into issues in the past when
moving files from OpenVMS to Unix/Mac/Windows. All those "pesky"
OpenVMS file attributes that sometimes get "corrupted".

If I am storing Mac files on OpenVMS I don't see a problem.

But if I move a file from OpenVMS (save a backup saveset), and then
move it back to the same or different OpenVMS system from the Mac,
will there be any "problems" you can think of??

John E. Malmberg

unread,
Mar 14, 2011, 12:15:21 AM3/14/11
to

I do not know how anything other than HP TCPIP does it.

Some of this is controlled by the mount options.

The VMS attributes for the file are put in a second "hidden" file with a
name that is based on the file name.

I have seen two issues with using NFS files on VMS served from Cygwin NFS.

The VMS debugger, at least the Motif version of it will hang if the
source files are on an NFS disk.

The DCL search command also hangs when searching files on the NFS disk.

The only way to clear the hang appears to be to stop the hung process.

-John
wb8...@qsl.network
Personal Opinion Only

JF Mezei

unread,
Mar 14, 2011, 12:58:11 AM3/14/11
to
rcyoung wrote:
> A quick "related" question- I have run into issues in the past when
> moving files from OpenVMS to Unix/Mac/Windows. All those "pesky"
> OpenVMS file attributes that sometimes get "corrupted".

Most reliable way I have found is to ZIP the files on VMS and unzip them
on the mac. This way text files take on native Unix attributes, at which
point VMS can access them as stream-lf files.

VAXman-

unread,
Mar 14, 2011, 9:49:00 AM3/14/11
to


On your Mac, in your $HOME directory, issue either (examples with me logged
in as VAXman):

% pwd
/Users/VAXman

...or...

% echo $HOME
/Users/VAXman

The /Users in the /etc/export refers to the /Users directory where Max OS X
builds all user's account home directories. The manpage 'man exports' says
this:

The first field(s) of an export entry is a list of directories on a local
file system to export. At least one pathname must be to an exported
directory. Other pathnames may refer to sub-directories of the exported
directory to indicate that hosts are also allowed to explicitly mount
those sub-directories of the exported directory. (Note that the -alldirs
flag can be used to allow mounting any sub-directories of the pathnames
specified.)

VAXman-

unread,
Mar 14, 2011, 9:57:22 AM3/14/11
to
In article <a36366f9-73b5-4d0f...@r19g2000prm.googlegroups.com>, rcyoung <rcy...@aliconsultants.com> writes:
>I am (at least a partial) idiot. I wondered what "/Users was"...I
>should have reread the posting to see he was talking about the
>"contents" of the /etc/export file.

Maybe I presumed you were more than a novice with nfs. Perhaps, I should
have explained that with greater detail.


>Just want to be sure I "catch
>myself" on that one before everyone thinks I'm a dolt.....it's been a
>LONG DAY :-)

As long as you're not a faineant dolt and you learn something, I wouldn't
think you a dolt.

Bob Koehler

unread,
Mar 14, 2011, 11:07:26 AM3/14/11
to
> Has anyone had experience with getting NFS file sharing to/from
> OpenVMS (8.2) from/to the Mac (10.6.6)? I can get the Mac volume
> mounted over on OpenVMS, but any attempt to do a "dir" or "copy"
> results in a timeout. Here is the terminal output

I haven't done this with HP's IP stack, but I have done it with
Multinet.

The one issue I ran into was that Multinet was using an older NFS
protocol and the Mac wasn't failing over to it. When I forced the
Mac to use the older protocol, it ran fine.

Bob Koehler

unread,
Mar 14, 2011, 11:09:50 AM3/14/11
to

Yes. Sadly, even Mac uses a brain-dead file system and can't
properly track file attributes. Various FTP clients and servers
have workarounds for this, but not NFS.

Richard B. Gilbert

unread,
Mar 14, 2011, 10:38:45 AM3/14/11
to


WHAT PROBLEM ARE YOU TRYING TO SOLVE??????????????????????????????


I can't think of any "problem" but I would not want to experiment
without first making backups of all disks that I would be writing to.
If I had important stuff on disks on this system I would want to make
backups of EVERYTHING.

rcyoung

unread,
Mar 14, 2011, 12:28:47 PM3/14/11
to
OK- I have NFS working under OpenVMS 8.2 to Mac 10.6.6 (bottom
posting). They display a bit "funky" but "any port in a storm"........
Now let's time travel a bit...to OpenVMS 7.1 and UCX. I have tried
setting up the same thing on the older system, and have a glitch in
that the NFS client is not creating the mount point automatically
(screen immediately follows) . Using same settings as before (except
VMS username is different), and I am only using the VMS NFS client in
BOTH instances

UCX> show proxy

VMS User_name Type User_ID Group_ID Host_name

FREND ON 501 501 *
UCX> show hosts

LOCAL database

Host address Host name

127.0.0.1 LOCALHOST, localhost
208.246.88.242 lurker2, LURKER2
208.246.88.161 rcyoung, RCYOUNG
UCX>


UCX> mount DNFS0: /HOST="rcyoung" /path="/Users/rcyoung" /
automount=inactivity:2
3:00:00/force
%UCX$DNFSMOUNT-E-MOUNTFAIL, error mounting DNFS0:
-SYSTEM-W-NOSUCHDEV, no such device available
UCX>

Been looking around for the old UCX manual "online" in case there is a
"missing bit", Do you see anything?

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

HURRAY it works on OpenVMS 8.2 (see below)- one down


$ sh dev d

Device Device Error Volume Free
Trans Mnt
Name Status Count Label Blocks
Count Cnt
A2100N$DKA0: Online 0
A2100N$DKA100: Mounted 0 ALPHA820

844200 348 1


A2100N$DKA200: Mounted 0 AXPDATA2
796608 1 1
A2100N$DKA300: Mounted 0 AXPDATA3
997408 1 1
A2100N$DKA400: Online wrtlck 0
A2100N$DVA0: Online 0
DNFS0: Online 0

DNFS17: Mounted 0 rcyou$/Users
********* 1 1
$ dir dnfs17:[000000]

Directory DNFS17:[000000]

$5N$M$ERIDIAN3.DIR;1
1/1 6-AUG-2010 09:30:56.00
$5N$M$ILLERS$R.DIR;1
1/1 12-OCT-2010 13:31:08.00
$5N$P$RO$MPIX.DIR;1
1/1 23-OCT-2009 18:10:44.00
$5N$ROES.DIR;1 1/1 5-MAY-2010 15:15:43.00
$5N$S$OFT$W$ORKS.DIR;1
1/1 25-OCT-2009 08:13:29.00
$5N$T$RASH.DIR;1 18/18 13-MAR-2011 15:04:19.00
$5N$X$CODE.DIR;1 1/1 20-JAN-2008 13:16:54.00
$5NADOBE.DIR;1 1/1 4-MAY-2010 14:24:55.00
$5NANTIWORD.DIR;1 2/2 6-AUG-2007 21:42:03.00
$5NCACHE.DIR;1 1/1 16-NOV-2009 10:58:12.00
$5NCONFIG.DIR;1 1/1 16-NOV-2009 11:11:46.00
$5NCUPS.DIR;1 1/1 29-OCT-2007 09:37:22.00
$5NDVDCSS.DIR;1 1/1 17-AUG-2008 09:51:53.00
$5NFONTCONFIG.DIR;1
10/10 6-FEB-2011 23:36:05.00
$5NJUNIQUE.DIR;1 1/1 6-FEB-2011 23:35:47.00
$5NLOCAL.DIR;1 1/1 16-NOV-2009 10:58:12.00
$5NMONO.DIR;1 1/1 5-FEB-2011 14:08:10.00
$5NMPLAYER.DIR;1 1/1 11-AUG-2008 22:00:53.00
$5NPARALLELS.DIR;1 1/1 31-DEC-2009 10:47:01.00
$5NROESCACHE.DIR;1 130/130 6-FEB-2011 23:33:55.00
$5NSHSH.DIR;1 3/3 28-NOV-2010 09:05:43.00
$5NSMB.DIR;1 1/1 16-AUG-2009 13:31:18.00
$5NSPUMUX.DIR;1 2/2 11-AUG-2008 21:59:17.00
$5NSSH.DIR;1 1/1 22-DEC-2008 10:10:37.00
$5NTHUMBNAILS.DIR;1
1/1 16-NOV-2009 11:00:16.00
$5NTOR.DIR;1 1/1 31-DEC-2010 23:36:25.00
$5NWIRESHARK-ETC.DIR;1
1/1 12-JAN-2011 08:47:36.00
$5NWIRESHARK.DIR;1 1/1 3-MAY-2009 17:35:40.00
$A$PPLICATIONS.DIR;1
1/1 4-MAY-2010 15:36:49.00
$ALI_$TECH_SUPPORT.TXT;1
0/0 9-APR-2009 21:44:39.00
$D$ESKTOP.DIR;1 9/9 13-MAR-2011 14:14:38.00
$D$OCUMENTS.DIR;1 17/17 7-MAR-2011 23:54:10.00
$D$OWNLOADS.DIR;1 3/3 13-MAR-2011 16:15:31.00
$DBT$UNE.DIR;1 1/1 12-MAR-2011 21:11:19.00
$L$IBRARY.DIR;1 4/4 13-MAR-2011 07:25:06.00
$M$OVIES.DIR;1 1/1 24-FEB-2011 16:53:36.00
$M$USIC.DIR;1 1/1 21-FEB-2009 21:46:47.00
$P$DA$N$ET$U$NINSTALL.SH;1
1/1 14-DEC-2010 23:08:38.00
$P$ICTURES.DIR;1 2/2 7-MAR-2011 23:52:59.00
$P$UBLIC.DIR;1 1/1 13-MAR-2011 15:54:00.00
$S$ITES.DIR;1 1/1 5-MAY-2010 13:19:05.00
$VMS821I_SYS-V0800.$TXT;1
38/38 15-DEC-2008 19:31:25.00
$VMS821I_SYS-V0800.ZIPEXE;1
11806/11806 15-DEC-2008 19:31:25.00
$Z$INIO$7A$L$IBRARY.DIR;1
1/1 18-DEC-2008 01:54:45.00
.$CFU$SER$T$EXT$E$NCODING;1
1/1 10-OCT-2007 10:44:58.00
.$DS_S$TORE;1 49/49 13-MAR-2011 16:23:51.00
.$X$AUTHORITY;1 1/1 6-FEB-2011 23:36:31.00
.BASH_HISTORY;1 15/15 13-MAR-2011 14:40:17.00
.JUPLOAD$5NPROPERTIES;1
1/1 3-OCT-2009 16:39:03.00
.LESSHST;1 1/1 10-AUG-2009 13:20:12.00
.NEW_CONTACT_$ALI$5N$TXT$5NSWP;1
8/8 9-APR-2009 21:45:08.00
.PARALLELS_SETTINGS;1
1/1 27-NOV-2008 19:23:35.00
.RECENTLY-USED$5NXBEL;1
3/3 16-NOV-2009 11:12:51.00
.TEST1$5NSWO;1 8/8 5-FEB-2008 16:17:50.00
.TEST1$5NSWP;1 8/8 5-FEB-2008 16:17:20.00
.VIMINFO;1 7/7 22-JAN-2011 04:48:47.00
.VUESCANRC;1 1/1 1-SEP-2007 10:44:16.00
LBX_FILES.DIR;1 6/6 11-MAR-2011 10:03:52.00
MBOX.;1 77/77 22-JUL-2009 21:41:38.00
NCO4.ZIP;1 187/187 13-FEB-2011 16:16:11.00
NEW_CONTACT_$ALI.$TXT;1
0/0 9-APR-2009 21:45:15.00
RESCUEPRO.PROPERTIES;1
1/1 31-DEC-2009 21:16:09.00
RESCUEPRO34ACT.LIC;1
2/2 31-DEC-2009 21:56:38.00

Total of 63 files, 12451/12451 blocks.
$


rcyoung

unread,
Mar 14, 2011, 12:31:53 PM3/14/11
to
I'm just thinking ahead to look at all the possibilities/options at
this time.

VAXman-

unread,
Mar 14, 2011, 6:09:38 PM3/14/11
to
In article <0bf6982a-6fcd-4fbe...@17g2000prr.googlegroups.com>, rcyoung <rcy...@aliconsultants.com> writes:
>OK- I have NFS working under OpenVMS 8.2 to Mac 10.6.6 (bottom
>posting). They display a bit "funky" but "any port in a storm"........

You need to add /structure=5 to your TCPIP> MOUNT command to get
this mounted as an ODS5 volume.

>Now let's time travel a bit...to OpenVMS 7.1 and UCX. I have tried
>setting up the same thing on the older system, and have a glitch in
>that the NFS client is not creating the mount point automatically
>(screen immediately follows) . Using same settings as before (except
>VMS username is different), and I am only using the VMS NFS client in
>BOTH instances

Yikes... back in those days, I used MultiNet. I'll defer to somebody
else to answer that for you.

rcyoung

unread,
Mar 14, 2011, 5:37:33 PM3/14/11
to
I remember Multinet. Worked very well.

rcyoung

unread,
Mar 18, 2011, 10:53:09 PM3/18/11
to
Have things working for NFS under 7,1 as well ...so far so good.

Now another issue....NSF keeps the VMS file characteristics in
"invisible" ADF files created along with the OpenVMS files on the NSF
server...so far so good.

If one were to try a ZIP from the VMS side, of a file stored on a NSF
volume, would it "automatically" incorporate the file characteristics,
or would they be "lost" ? Remember they are kept in a separate file
under NSF, but does the OS "automatically" bring them in when the
files are "accessed" and zipped???


Don't I come up with the most "interesting" questions?????


JF Mezei

unread,
Mar 18, 2011, 11:53:56 PM3/18/11
to
rcyoung wrote:

> If one were to try a ZIP from the VMS side, of a file stored on a NSF
> volume, would it "automatically" incorporate the file characteristics,
> or would they be "lost" ? Remember they are kept in a separate file
> under NSF, but does the OS "automatically" bring them in when the
> files are "accessed" and zipped???


I *think* there are settings on the VMS side on whether attributes are
saved or not. I have found NFS to be unreliable for large text files.


Here is an example:

> velo:DMA2 $ ls

> .$ADF$A1032_IGUIDE.PS;1 .Spotlight-V100 A1032_IGUIDE.PS Backup NETGEAR.ZIP vms_alpha_1of9.cdr
> .$ADF$A1032_IGUIDE.PS;2 .TemporaryItems ALLIN1.SAV
> .$ADF$ALLIN1.SAV;1

From VMS:

> dir/full dnfs2:[000000]allin1.sav
>
> Directory DNFS2:[000000]
>
> ALLIN1.SAV;1 File ID: (1822,31656,0)
> Size: 223650/223650 Owner: [SYSTEM]
> Created: 5-MAR-2011 02:33:30.00
> Revised: 5-MAR-2011 03:05:22.00 (1)
> Expires: <None specified>
> Backup: <No backup recorded>
> Effective: <None specified>
> Recording: <None specified>
> Accessed: 5-MAR-2011 03:29:59.00
> Attributes: 5-MAR-2011 03:05:22.00
> Modified: 31-DEC-1969 20:00:00.00
> Linkcount: 1
> File organization: Sequential
> Shelved state: Online
> Caching attribute: No_caching
> File attributes: Allocation: 223650, Extend: 0, Global buffer count: 0
> No version limit
> Record format: Fixed length 32256 byte records
> Record attributes: None
> RMS attributes: None
> Journaling enabled: None
> File protection: System:RWED, Owner:RWED, Group:RE, World:
> Access Cntrl List: None
> Client attributes: None
>
> Total of 1 file, 223650/223650 blocks.

> $ backup/list dnfs2:[000000]allin1.sav/save
> Listing of save set(s)
>
> Save set: ALLIN1.SAV
> Written by: SYSTEM
> UIC: [000001,000004]
> Date: 5-MAR-2011 01:33:30.67
> Command: BACKUP/IGNORE=INTERLOCK $DISK2:[ALLIN1...]*.* DNFS2:[000000]ALLIN1.SAV/SAVE/LOG
> Operating system: OpenVMS Alpha version V8.3
> BACKUP version: V8.3
> CPU ID register: 80000000
> Node name: _CHAIN::
> Written on: _DNFS2:
> Block size: 32256
> Group size: 10
> Buffer count: 198
>
> [ALLIN1]ADMIN_DATA.DIR;1 2 11-JAN-2007 00:36
> [ALLIN1.ADMIN_DATA]ZBOR7GD1X.DAT;1 6 18-FEB-2006 05:13
> [ALLIN1.ADMIN_DATA]ZBOR7IXNC.DAT;1 6 18-FEB-2006 05:16
> [ALLIN1.ADMIN_DATA]ZBOY8MEH3.DAT;1 6 25-FEB-2006 06:02
> [ALLIN1.ADMIN_DATA]ZBUIKV1GY.DAT;1 6 15-JUL-2006 14:36
> [ALLIN1.ADMIN_DATA]ZUHZTO9NQ.DAT;1 9 13-SEP-1992 20:46
> [ALLIN1.ADMIN_DATA]ZUIIWFP8Z.DAT;1 9 22-SEP-1992 22:42
> [ALLIN1.ADMIN_DATA]ZUIIWHEZ8.DAT;1 9 22-SEP-1992 22:44
> [ALLIN1.ADMIN_DATA]ZULQWUA30.DAT;1 9 17-DEC-1992 22:59
> [ALLIN1.ADMIN_DATA]ZULQWUA31.DAT;1 15 17-DEC-1992 23:00
> [ALLIN1.ADMIN_DATA]ZULQWUA32.DAT;1 15 17-DEC-1992 23:00
> [ALLIN1.ADMIN_DATA]ZUQPU2THE.DAT;1 9 25-APR-1993 21:03

So it appears to work

rcyoung

unread,
Mar 19, 2011, 7:28:21 PM3/19/11
to
I can see what you mean...I am trying to see if I can make a image
backup saveset on the NFS disk of various hard drives. THe 2-4Gb disks
seem to work. But the 32Gb RAID array
image backup "dies"...the backup command executes, but the saveset
file gets to around 4 Gb and then things go crazy, and in the in the
saveset file is suddenly < 2Gb in size...yet looking at the log on VMS
shows everything seemed to be fine. So the backup command "thinks" it
is ok...but something in the NFS is killing it.

Richard B. Gilbert

unread,
Mar 19, 2011, 8:28:45 PM3/19/11
to

Are you talking about VAX, Alpha, or Itanic? I would not be totally
surprised if a VAX couldn't do it. The Alpha and Itanic architectures
should have no problem.

rcyoung

unread,
Mar 20, 2011, 9:56:42 AM3/20/11
to
> Are you  talking about VAX, Alpha, or Itanic?  I would not be totally
> surprised if a VAX couldn't do it.  The Alpha and Itanic architectures
> should have no problem.

Alpha----but I have seen posts going back several years where people
"alluded" to, but did not provide a lot of info, on NFS probs with
large files. Maybe this is why?

rcyoung

unread,
Mar 20, 2011, 10:21:12 AM3/20/11
to

Here is what I got....the backup log starts with


Save set: LURKER_NFS_DRA0.
Written by: { blanked}
UIC: { blanked}
Date: 19-MAR-2011 07:29:38.45
Command: BACKUP/IMAGE/RECORD/IGNORE=(INTERLOCK)/LIST=DRA0:
[LOGS]LURKER
_NFS_DRA0.LOG DRA0: DNFS1:[000000]LURKER_NFS_DRA0/SAV
Operating system: OpenVMS AXP version V7.1
BACKUP version: V6.2
CPU ID register: 80000000
Node name: { blanked}::
Written on: _$1$DNFS1:


Block size: 32256
Group size: 10

Buffer count: 31

Image save of volume set
Number of volumes: 1

Volume attributes
Structure level: 2
Label: ALPHASYS

Owner:
Owner UIC: [000001,000001]
Creation date: 27-SEP-2001 15:00:02.70
Total blocks: 33513472
<--------------------------------------------
Access count: 3
Cluster size: 33
Data check: No Read, No Write
Extension size: 5


File protection: System:RWED, Owner:RWED, Group:RE, World:

Maximum files: 492845
Volume protection: System:RWCD, Owner:RWCD, Group:RWCD, World:RWCD

*
*
*
Total of 22730 files, 26768611 blocks
<-------------------------------------------- what the backup/image
thinks it did at end of log
End of save set

Backup/list of resulting file


$ backup/list dnfs1:[000000]lurker_nfs_dra0.;1/sav
<-------------------------------------------- error you get
Listing of save set(s)

%BACKUP-E-READERRS, excessive error rate reading DNFS1:
[000000]LURKER_NFS_DRA0.;
1
-BACKUP-E-HDRCRC, software header CRC error
%BACKUP-I-OPERSPEC
%BACKUP-I-OPERASSIST, operator assistance has been requested


And a "ls -l" from the UNIX side of the saveset file

-rwxr-x--- 1 rcyoung rcyoung 1661457408 Mar 19 16:48
lurker_nfs_dra0 <-------------------------------------------- final
size, & as I was monitoring it during the backup, it "was" over 4 Gb,
then "poof"

Wilm Boerhout

unread,
Mar 20, 2011, 1:16:26 PM3/20/11
to
rcyoung mentioned on 20-3-2011 15:21:
> size,& as I was monitoring it during the backup, it "was" over 4 Gb,
> then "poof"
>

Your first post indicated VMS V8.2, this now looks as it's a much older
version. I know from experience that VMS $BACKUP had no possibility to
write tape savesets of larger than 4GB on that version. ISTR that the
error messages were the same as above. Maybe the same restriction
applies to disk save sets?

--
Wilm Boerhout |
Zwolle, | 'These are the times that try men's souls'
The Netherlands | (Thomas Paine, The Crisis)

JF Mezei

unread,
Mar 20, 2011, 4:42:29 PM3/20/11
to
rcyoung wrote:

> Alpha----but I have seen posts going back several years where people
> "alluded" to, but did not provide a lot of info, on NFS probs with
> large files. Maybe this is why?
>

I believe that older versions of NFS support only 32 bits in the file
size. This would explain the 4 gig limit.

rcyoung

unread,
Mar 21, 2011, 6:17:38 AM3/21/11
to
Perhaps more experimenting will tell..Yes, I am working on a mix of
7.1 thru 8.2 VMS systems. Wouldn't it be nice to find (1) solution
that worked across the board? Oh well, time will tell.

Paul Sture

unread,
Mar 27, 2011, 8:24:21 AM3/27/11
to

This is the problem I mentiond in my other post. Above 2GB (or was
it 4GB?) the saveset contents "wrapped" to the beginning. It was easy
enough to see when I did a DUMP of the saveset.

0 new messages