RE: (no subject)

RE: (no subject)

[Lever, Charles]

hi steve-

the limitation you point out below is real, and would be
addressed by Kerberos authentication.  however...

Linux NFS does not support RPCSEC mechanisms in the stable
kernel at the moment.  that support is going into the
development kernel (2.5) over the next few months, and
will be available in a stable kernel when 2.6 appears, at
least for NFSv4.  if we're all very very good, RPCSEC may
also be supported for the older versions of NFS too.

> -----Original Message-----
> From: Steve Salazar [mailto:eagsalazar@hotmail.com]
> Sent: Sunday, March 09, 2003 2:58 PM
> To: nfs@lists.sourceforge.net
> Subject: [NFS] (no subject)
>=20
>=20
> I have been reading the nfs documentation and mailing lists=20
> and it is not=20
> clear to me whether the current nfs implementation supports kerberos=20
> authentication or not.  There is nothing about that in the=20
> howto but in the=20
> mailing list I see repeated references to "NFS using=20
> RPCSEC_GSS".  However,=20
> I haven't found any documentation on this either.
>=20
> This issue mentioned in the howto:
>=20
> "But the root user on the client can still use su to become=20
> any other user=20
> and access and change that users files!" say you. To which=20
> the answer is:=20
> Yes, and that's the way it is, and has to be with Unix and=20
> NFS. This has one=20
> important implication: All important binaries and files=20
> should be owned by=20
> root, and not bin or other non-root account, since the only=20
> account the=20
> clients root user cannot access is the servers root account.
>=20
> will not work for for our setup.  We have kerberos in place=20
> for logins on=20
> our linux/solaris/windows network and a kerberized samba but=20
> so far we have=20
> no good solution for nfs since we will definitely need to=20
> allow local root=20
> on the linux workstations on our network.
>=20
> Could anyone point me at some good documentation on how to=20
> deploy nfs using=20
> kerberos?  Thanks in advance for any info.
>=20
>=20
>=20
> _________________________________________________________________
> The new MSN 8: advanced junk mail protection and 2 months FREE* =20
> http://join.msn.com/?page=3Dfeatures/junkmail
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Etnus, makers of=20
> TotalView, The debugger=20
> for complex code. Debugging C/C++ programs can leave you=20
> feeling lost and=20
> disoriented. TotalView can help you find your way. Available=20
> on major UNIX=20
> and Linux platforms. Try it free. www.etnus.com
> _______________________________________________
> NFS maillist  -  NFS@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/nfs
>=20


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

NFS Benchmarking tests?

[Brent M. Clements]

What are some good NFS Benchmarking tests that I could run.

I already know the generic answers, such as use iozone, or use this or
that...Could I get some specifics? I need to test the throughput of 1
node, 16 nodes, 32 nodes and so on and so on.

Thanks,
Brent





-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: NFS Benchmarking tests?

[Steve Dickson]

Try http://www.cs.duke.edu/ari/fstress/

It seems to do a good job on server benchmarking.

SteveD.

Brent M. Clements wrote:

>What are some good NFS Benchmarking tests that I could run.
>
>I already know the generic answers, such as use iozone, or use this or
>that...Could I get some specifics? I need to test the throughput of 1
>node, 16 nodes, 32 nodes and so on and so on.
>
>Thanks,
>Brent
>
>
>
>
>
>-------------------------------------------------------
>This sf.net email is sponsored by:ThinkGeek
>Welcome to geek heaven.
>http://thinkgeek.com/sf
>_______________________________________________
>NFS maillist  -  NFS@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/nfs
>  
>



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: NFS Benchmarking tests?

[Wendy Cheng]

Not sure how "specific" you like but SPECsfs97 
has been adopted by many big name vendors:

http://www.spec.org/sfs97r1/

Wendy
 
----- Original Message ----- 

| Date: Mon, 10 Mar 2003 14:06:45 -0600 (CST)
| From: "Brent M. Clements" <bclem@rice.edu>
| To: <nfs@lists.sourceforge.net>
| Subject: [NFS] NFS Benchmarking tests?
| 
| What are some good NFS Benchmarking tests that I could run.
| 
| I already know the generic answers, such as use iozone, or use this or
| that...Could I get some specifics? I need to test the throughput of 1
| node, 16 nodes, 32 nodes and so on and so on.
| 
| Thanks,
| Brent


-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en
_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs