NFS Error

[Willem Grooters]

Peter,

Thinks like this have been discussed before:

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=448989

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=246673

Perhaps your answer is in one of these

Willem

[Bojan Nemec]

Peter,

Try to modify yours /etc/exports line to:

/data/vmsalpha(sync) 129.100.2.1/16(rw)

Try as Faris suggested:

su - nfs_user -c "echo test > /data/vmsalpha/test"
ls -l /data/vmsalpha/test

If the owner of /data/vmsalpha is not nfs_user do a:

chown 400:400 /data/vmsalpha

By the way (sorry if this is a stupid question but sometimes happens) is the directory /data/vmsalpha created?

Bojan

[Peter Clarke]

I can now set my default as dnfs1:[000000] and see the files in this directory.
However when i try to write a file it gives the message:

Device currently write locked

Reg

Peter

[Bojan Nemec]

Peter,

What is the exact reported error? Can you try to access to this NFS disk from another Linux machine?

Bojan

[Peter Clarke]

Exact error is:

-RMS-E-WLK, device currently write locked

Unfortunately i don't have another Linux machine!

Peter

[faris_3]

can you temporarily do
a chmod 777 on the directory on linux
(access to everybody)
to see if that works

[Bojan Nemec · Accepted solution]

Peter,

The only reported error in VMS abbout RMS-E-WLK is taht you have mounted the NFS device whit the /NOWRITE qualifier. I suppose that you have not done this.

So probably there is an error on the Linux side. Please see the Linux NFS-HOWTO
http://www.faqs.org/docs/Linux-HOWTO/NFS-HOWTO.html

maybe the most interesant for you is
http://www.faqs.org/docs/Linux-HOWTO/NFS-HOWTO.html#SYMPTOM4

Bojan

[Bojan Nemec]

Maybe this page will also be a good reading:

http://nfs.sourceforge.net/

Bojan

[Peter Clarke]

ok,I set my oracle dump file to save to this device (dnsf1:) and checked it today and the file has been output to this device and freed up lots of space on my alpha.

However the dump file on the Alpha was over 14GB but the one on the Linux is only 1.41GB i have checked the log file and it seems to have copied all the data but im not sure it has done this correctly!! and have no real way of checking this dump file ie. no spare machine and not enough space to create a new DB on the alpha or linux box.

Also the oracle export normally takes 1 and a half hours but to the Linux box it took 5 and a half is this normal? any way of improving this??

Cheers

Peter

[Ian Miller.]

as you are now exporting your database over a network connection rather than a directly connected disk I would expect it to be slower.

As the file sizes you mentioned are a factor of 10 different could there be an arithmentic error in how you are obtaining the file sizes?