Discussion:
Problem with NFS
(too old to reply)
Joukj
2017-06-13 06:32:59 UTC
Permalink
Raw Message
Hi all,

I got a problem writing from my OpenVMS on a NFS share locxated on a
Fedora 25 system. My TCP/IP stack is the following:

HP TCP/IP Services for OpenVMS Alpha Version V5.7 - ECO 5
on a COMPAQ Professional Workstation XP1000 running OpenVMS V8.4




I can mount the share :

$ tcpip mount dnfs0: FOXTROT$backup FOXTROT$backup
/host=foxtrot.nano.tudelft.nl/path="/backup"/system/uid=0/gid=0/acp=(max_work=1024)/timeout=::30
%TCPIP$DNFSMOUNT-S-MOUNTED, /backup mounted on _DNFS2:[000000]





I can create a file on the share:

$ set def FOXTROT$backup:[joukj.vms_backup]
$ create tmp.tmp
sadsafsd
Exit
troika-jj) dir tmp

Directory FOXTROT$BACKUP:[JOUKJ.VMS_BACKUP]

TMP.TMP;1

Total of 1 file






But when I do this I get:

$ edit/edt tmp.tmp
%EDT-F-OPENOUT, error opening FOXTROT$BACKUP:[JOUKJ.VMS_BACKUP]TMP.JOU;
as output
-RMS-F-ATW, file attributes write error
-SYSTEM-W-FCPWRITERR, file processor write error

.
Why do I get this and how can I solve this?





Regards

Jouk
Volker Halle
2017-06-13 06:43:33 UTC
Permalink
Raw Message
Hi Jouk,

did you try using TCPIP MOUNT/ADF ?

Volker.
Joukj
2017-06-13 07:09:50 UTC
Permalink
Raw Message
Post by Volker Halle
Hi Jouk,
did you try using TCPIP MOUNT/ADF ?
Volker.
Tried but did not help.

Jouk
Volker Halle
2017-06-13 07:21:22 UTC
Permalink
Raw Message
Jouk,

you may need a TCPIP patch: TCPIP_NFS_PAT-V0507-ECO5C

See https://vmssoftware.com/pdfs/VSI_OpenVMS_V842L1_CLRN.pdf

The patch should also be available from HPE.

Volker.
Joukj
2017-06-13 07:37:02 UTC
Permalink
Raw Message
Post by Volker Halle
Jouk,
you may need a TCPIP patch: TCPIP_NFS_PAT-V0507-ECO5C
See https://vmssoftware.com/pdfs/VSI_OpenVMS_V842L1_CLRN.pdf
The patch should also be available from HPE.
Volker.
The whole thing may be triggered by an upgrade of the fedora nfs server,,,

sadly, since the drop of the alliance-one program, I lost my access to
the patches.

regards
Jouk
Scott Dorsey
2017-06-13 10:27:42 UTC
Permalink
Raw Message
Post by Joukj
The whole thing may be triggered by an upgrade of the fedora nfs server,,,
You do know that there was a recent red hat patch that totally breaks RPC
and all RPC-related service, right? It can be backed out although it's been
a couple days now and there should be a patch for the patch.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
John E. Malmberg
2017-06-13 12:45:52 UTC
Permalink
Raw Message
Post by Joukj
Post by Volker Halle
Jouk,
you may need a TCPIP patch: TCPIP_NFS_PAT-V0507-ECO5C
See https://vmssoftware.com/pdfs/VSI_OpenVMS_V842L1_CLRN.pdf
The patch should also be available from HPE.
Volker.
The whole thing may be triggered by an upgrade of the fedora nfs server,,,
Use /noadf if you can.

https://bugzilla.redhat.com/show_bug.cgi?id=1460255
Post by Joukj
sadly, since the drop of the alliance-one program, I lost my access to
the patches.
I no longer have access to patches either.

I inquired to HPE about access to the patch on Friday. No response so far.

Regards,
-John
***@qsl.net_work
Joukj
2017-06-14 06:13:48 UTC
Permalink
Raw Message
Post by John E. Malmberg
Post by Joukj
Post by Volker Halle
Jouk,
you may need a TCPIP patch: TCPIP_NFS_PAT-V0507-ECO5C
See https://vmssoftware.com/pdfs/VSI_OpenVMS_V842L1_CLRN.pdf
The patch should also be available from HPE.
Volker.
The whole thing may be triggered by an upgrade of the fedora nfs server,,,
Use /noadf if you can.
https://bugzilla.redhat.com/show_bug.cgi?id=1460255
Post by Joukj
sadly, since the drop of the alliance-one program, I lost my access to
the patches.
I no longer have access to patches either.
I inquired to HPE about access to the patch on Friday. No response so far.
Regards,
-John
Thanks John

Please let me know when and how HPE responds. When I have time I will
also try to contact them.

I;m not sure if /noadf will work for me, since I use the NFS share to
make backups (save-sets) of my disks to store them in a different
physixal location.

I will add also a note to the bug report at redhat (i realy do not like
the last remark in it....)

Regards
Jouk
John E. Malmberg
2017-06-14 12:32:38 UTC
Permalink
Raw Message
Post by Joukj
Post by John E. Malmberg
Post by Joukj
The whole thing may be triggered by an upgrade of the fedora nfs server,,,
Use /noadf if you can.
https://bugzilla.redhat.com/show_bug.cgi?id=1460255
I inquired to HPE about access to the patch on Friday. No response so far.
Please let me know when and how HPE responds. When I have time I will
also try to contact them.
I;m not sure if /noadf will work for me, since I use the NFS share to
make backups (save-sets) of my disks to store them in a different
physixal location.
The 8.4 VMS backup should be able to handle that. Older VMS need the
fix_saveset.com available on the wild wild web.
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not like
the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.

I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.

I see no malice in the last remark, or need to respond to it there.

That is just showing that VSI has to find a way to reverse that
perception in the IT world.

Regards,
-John
***@qsl.net_work
John E. Malmberg
2017-06-14 12:43:53 UTC
Permalink
Raw Message
Post by John E. Malmberg
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not
like the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.
I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.
I see no malice in the last remark, or need to respond to it there.
Just saw your comment, and I think it is a appropriate. I am sorry if I
jumped to the wrong conclusion from your earlier remark.

Regards,
-John
***@qsl.net_work.
Simon Clubley
2017-06-14 18:10:51 UTC
Permalink
Raw Message
Post by John E. Malmberg
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not like
the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.
I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.
I see no malice in the last remark, or need to respond to it there.
There is indeed no malice there.

VMS is now so far out of the mindset of much of the general computing
community that reporting an OpenVMS issue to them would be like saying
to one of us here that something breaks when it's run on a PDP-11
operating system and then asking for a fix.

You would get the same reaction here - that surely those PDP-11
operating systems are now dead. Well, that's how most of the general
computing community views VMS.
Post by John E. Malmberg
That is just showing that VSI has to find a way to reverse that
perception in the IT world.
Agreed. And so far, I am not seeing any real public advocacy that
demonstrates they are doing this.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
David Froble
2017-06-14 18:58:36 UTC
Permalink
Raw Message
Post by Simon Clubley
Post by John E. Malmberg
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not like
the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.
I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.
I see no malice in the last remark, or need to respond to it there.
There is indeed no malice there.
VMS is now so far out of the mindset of much of the general computing
community that reporting an OpenVMS issue to them would be like saying
to one of us here that something breaks when it's run on a PDP-11
operating system and then asking for a fix.
You would get the same reaction here - that surely those PDP-11
operating systems are now dead. Well, that's how most of the general
computing community views VMS.
Perhaps I'm biased ..

Ok, I AM biased.

An attitude as described above just goes to show just how unprofessional such
people are. Not everything is going to be #1. Not being aware of other things
just shows ignorance. Ignorance of anything but weendoze, and perhaps *ux.

Then again, for those who think Bill gates invented computers, not much can be
said about such.

Got to wonder how many of such people know what an IBM Sysplex is? Or some of
the real time OSs available.
Post by Simon Clubley
Post by John E. Malmberg
That is just showing that VSI has to find a way to reverse that
perception in the IT world.
Agreed. And so far, I am not seeing any real public advocacy that
demonstrates they are doing this.
Timing is everything ....
Paul Sture
2017-06-14 20:51:21 UTC
Permalink
Raw Message
Post by Simon Clubley
Post by John E. Malmberg
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not like
the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.
I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.
I see no malice in the last remark, or need to respond to it there.
There is indeed no malice there.
VMS is now so far out of the mindset of much of the general computing
community that reporting an OpenVMS issue to them would be like saying
to one of us here that something breaks when it's run on a PDP-11
operating system and then asking for a fix.
You would get the same reaction here - that surely those PDP-11
operating systems are now dead. Well, that's how most of the general
computing community views VMS.
That presents an accurate picture of the reactions to VMS that I come
across. There are a surprising number of positive comments when VMS
does get a mention, but they are overwhelmingly of the "it was nice
when it was still popular" nature.
--
Everybody has a testing environment. Some people are lucky enough to
have a totally separate environment to run production in.
Joukj
2017-06-15 06:44:50 UTC
Permalink
Raw Message
Post by John E. Malmberg
Post by Joukj
Post by John E. Malmberg
Post by Joukj
The whole thing may be triggered by an upgrade of the fedora nfs server,,,
Use /noadf if you can.
https://bugzilla.redhat.com/show_bug.cgi?id=1460255
I inquired to HPE about access to the patch on Friday. No response so far.
Please let me know when and how HPE responds. When I have time I will
also try to contact them.
I;m not sure if /noadf will work for me, since I use the NFS share to
make backups (save-sets) of my disks to store them in a different
physixal location.
The 8.4 VMS backup should be able to handle that. Older VMS need the
fix_saveset.com available on the wild wild web.
Post by Joukj
I will add also a note to the bug report at redhat (i realy do not
like the last remark in it....)
HP declared VMS EOL, a non-VMS programmer would probably not know about VSI.
I am just surprised and grateful that RedHat responded almost
immediately to the bug and was willing to investigate network traces to
see what the issue was.
That is what I almost always see when I report a "bug" in Fedora: They
seem very eager to get a good product

Jouk

Loading...