Hi list, Any one has ANY new data on OpenSolaris vs Linux? I only found an old post in 2006. http://mail.opensolaris.org/pipermail/zfs-discuss/2006-January/030366.html And any comments on "if OpenSolaris performance is about the same as Solaris 10"? Thanks! z -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/a3d7c713/attachment.html>
On Wed, Dec 3, 2008 at 2:39 PM, Joseph Zhou <jz at excelsioritsolutions.com>wrote:> Hi list, > > Any one has ANY new data on OpenSolaris vs Linux? > > I only found an old post in 2006. > http://mail.opensolaris.org/pipermail/zfs-discuss/2006-January/030366.html > > And any comments on "if OpenSolaris performance is about the same as > Solaris 10"? > > Thanks! > z > >That''s kind of open ended. What sort of performance are you looking for? NFS throughput? Software raid? What distro vs. Solaris? Opensolaris and Solaris are going to have different performance based on what exactly it is you''re testing. Similar is probably accurate for a lot of things, but not everything. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/aedd6d36/attachment.html>
Thanks Tim,
At this moment, I am looking into OpenStorage as NAS (file serving) vs. Linux
NAS (Samba) vs. Win2008 NAS vs. NetApp (ONTAP, not GX) performance.
I am also interested in block-based performance, but not as urgent as above.
(Since 7000 is mainly doing NAS today, in a non-HPC-clustered fashion without
Lustre. With Lustre, the performance competitive focuses are different from
above).
Thanks,
z
----- Original Message -----
From: Tim
To: Joseph Zhou
Cc: zfs-discuss at opensolaris.org
Sent: Wednesday, December 03, 2008 4:04 PM
Subject: Re: [zfs-discuss] OpenSolaris vs Linux
On Wed, Dec 3, 2008 at 2:39 PM, Joseph Zhou <jz at
excelsioritsolutions.com> wrote:
Hi list,
Any one has ANY new data on OpenSolaris vs Linux?
I only found an old post in 2006.
http://mail.opensolaris.org/pipermail/zfs-discuss/2006-January/030366.html
And any comments on "if OpenSolaris performance is about the same as
Solaris 10"?
Thanks!
z
That''s kind of open ended. What sort of performance are you looking
for? NFS throughput? Software raid? What distro vs. Solaris?
Opensolaris and Solaris are going to have different performance based on what
exactly it is you''re testing. Similar is probably accurate for a lot
of things, but not everything.
--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/f50c2bd0/attachment.html>
Joseph Zhou wrote:> Thanks Tim, > At this moment, I am looking into OpenStorage as NAS (file serving) > vs. Linux NAS (Samba) vs. Win2008 NAS vs. NetApp (ONTAP, not GX) > performance. >There are still a number of ZFS/OpenSOlaris options to compare, iSCSI, Samba, CIFS, NFS. -- Ian.
On Wed, Dec 3, 2008 at 3:11 PM, Joseph Zhou <jz at excelsioritsolutions.com>wrote:> Thanks Tim, > At this moment, I am looking into OpenStorage as NAS (file serving) vs. > Linux NAS (Samba) vs. Win2008 NAS vs. NetApp (ONTAP, not GX) performance. > > I am also interested in block-based performance, but not as urgent as > above. (Since 7000 is mainly doing NAS today, in a non-HPC-clustered > fashion without Lustre. With Lustre, the performance competitive focuses are > different from above). > > Thanks, > z >Right, so hardware or software raid? NFS, CIFS, both? Win2k8 is going to blow serving NFS, but it can be done. Storage 7000 is going to have a COMPLETELY different performance envelope than vanilla opensolaris or solaris. With some customization using flash you might be able to get close, but if you want to know what a storage 7000 will do, you should ask for that, not just "opensolaris". Here''s an example of a loaded up 7000: http://blogs.sun.com/brendan/entry/a_quarter_million_nfs_iops If you want to compare it to something like NetApp though, it''s tough, because how do you make your comparison? Price? What model NetApp are you going to use? What kind of server are you going to use? If you just want to use some numbers someone comes up with to make a decision on what platform to use, I''d argue you''re going about it completely the wrong way. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/b088b3dc/attachment.html>
Thanks Ian, Tim, Ok, let me really hit one topic instead of trying to see in general what data are out there... Let''s say OpenSolaris doing Samba vs. Linux doing Samba, in CIFS performance. (so I can link to the Win2008 CIFS numbers and NetApp CIFS numbers myself.) Is there any data to this specific point? Thanks! z ----- Original Message ----- From: "Ian Collins" <ian at ianshome.com> To: "Joseph Zhou" <jz at excelsioritsolutions.com> Cc: "Tim" <tim at tcsac.net>; <zfs-discuss at opensolaris.org> Sent: Wednesday, December 03, 2008 4:31 PM Subject: Re: [zfs-discuss] OpenSolaris vs Linux> Joseph Zhou wrote: >> Thanks Tim, >> At this moment, I am looking into OpenStorage as NAS (file serving) >> vs. Linux NAS (Samba) vs. Win2008 NAS vs. NetApp (ONTAP, not GX) >> performance. >> > There are still a number of ZFS/OpenSOlaris options to compare, iSCSI, > Samba, CIFS, NFS. > > -- > Ian. >
On Wed, Dec 3, 2008 at 3:36 PM, Joseph Zhou <jz at excelsioritsolutions.com>wrote:> Thanks Ian, Tim, > Ok, let me really hit one topic instead of trying to see in general what > data are out there... > > Let''s say OpenSolaris doing Samba vs. Linux doing Samba, in CIFS > performance. > (so I can link to the Win2008 CIFS numbers and NetApp CIFS numbers myself.) > > Is there any data to this specific point? > Thanks! > z > >So, you wouldn''t use Samba on opensolaris, you''d use the native cifs stack. Then we have to look at the system itself. How much ram? How many and what kind of CPU''s? How much disk on the backend? What kind of disk on the back end? I don''t think you''re going to find the numbers you''re looking for to be quite honest. And even if you did, I don''t know how usable they''d really be. I''d start by digging through the spc benchmarks. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/d622d671/attachment.html>
Ok, thanks Tim, which SPC are you talking about?
SPC-1 and SPC-2 don''t test NAS, those are block perf.
SPECsfs97 v2/v3 and sfs2008 have no OpenStorage results.
If there are standard storage benchmarks out there, I would not be here asking
folks.
To your point, how is the OpenSolaris native CIFS vs Linux Samba then? (if you
think this is more apple-to-apple than they both run Samba)
Again, I am here to explore data, not to argue, if I give you a dozen
configurations, could you get me the performance estimates and how the estimates
come from? I didn''t think that route is possible.
Thanks.
z
----- Original Message -----
From: Tim
To: Joseph Zhou
Cc: Ian Collins ; zfs-discuss at opensolaris.org
Sent: Wednesday, December 03, 2008 4:40 PM
Subject: Re: [zfs-discuss] OpenSolaris vs Linux
On Wed, Dec 3, 2008 at 3:36 PM, Joseph Zhou <jz at
excelsioritsolutions.com> wrote:
Thanks Ian, Tim,
Ok, let me really hit one topic instead of trying to see in general what
data are out there...
Let''s say OpenSolaris doing Samba vs. Linux doing Samba, in CIFS
performance.
(so I can link to the Win2008 CIFS numbers and NetApp CIFS numbers myself.)
Is there any data to this specific point?
Thanks!
z
So, you wouldn''t use Samba on opensolaris, you''d use the
native cifs stack. Then we have to look at the system itself. How much ram?
How many and what kind of CPU''s? How much disk on the backend? What
kind of disk on the back end?
I don''t think you''re going to find the numbers
you''re looking for to be quite honest. And even if you did, I
don''t know how usable they''d really be. I''d start by
digging through the spc benchmarks.
--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/bf79d6ea/attachment.html>
On Wed, Dec 3, 2008 at 3:51 PM, Joseph Zhou <jz at excelsioritsolutions.com>wrote:> Ok, thanks Tim, which SPC are you talking about? > > SPC-1 and SPC-2 don''t test NAS, those are block perf. > SPECsfs97 v2/v3 and sfs2008 have no OpenStorage results. > > If there are standard storage benchmarks out there, I would not be here > asking folks. > > To your point, how is the OpenSolaris native CIFS vs Linux Samba then? (if > you think this is more apple-to-apple than they both run Samba) > > Again, I am here to explore data, not to argue, if I give you a dozen > configurations, could you get me the performance estimates and how the > estimates come from? I didn''t think that route is possible. > > Thanks. > z >Sorry, I was referring to SPEC, not SPC. Perhaps you could ask one of the folks from Sun on these mailing lists if they have plans to post results. I''d imagine they do for at least the storage 7000 series. I think native cifs on Solaris vs. Samba on Linux is fair simply because it''s what someone rolling out an implementation would use. It''ll never be 100% apples-to-apples, so I''d say real-world is preferred over hampering one system to make it *closer* to the other. As for configurations, I probably have access to enough hardware to do most of the benchmarking, but this time of year, being end-of-quarter, I wouldn''t have the time to do so. That doesn''t mean there isn''t someone else lurking who does. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/f847e704/attachment.html>
haha, Tim, yes, I see the Open spirit in this reply! ;-)
As I said, I am just exploring data.
The Sun J4000 SPC1 and SPC2 benchmark results were nice, just lacking other
published results with the iSCSI HBA as DAS, not as a network storage device (as
7000). Though I would attempt to say those results can be a basis for 7000
block-performance...
any comment?
Thanks!
z
----- Original Message -----
From: Tim
To: Joseph Zhou
Cc: Ian Collins ; zfs-discuss at opensolaris.org
Sent: Wednesday, December 03, 2008 5:00 PM
Subject: Re: [zfs-discuss] OpenSolaris vs Linux
On Wed, Dec 3, 2008 at 3:51 PM, Joseph Zhou <jz at
excelsioritsolutions.com> wrote:
Ok, thanks Tim, which SPC are you talking about?
SPC-1 and SPC-2 don''t test NAS, those are block perf.
SPECsfs97 v2/v3 and sfs2008 have no OpenStorage results.
If there are standard storage benchmarks out there, I would not be here
asking folks.
To your point, how is the OpenSolaris native CIFS vs Linux Samba then? (if
you think this is more apple-to-apple than they both run Samba)
Again, I am here to explore data, not to argue, if I give you a dozen
configurations, could you get me the performance estimates and how the estimates
come from? I didn''t think that route is possible.
Thanks.
z
Sorry, I was referring to SPEC, not SPC. Perhaps you could ask one of the
folks from Sun on these mailing lists if they have plans to post results.
I''d imagine they do for at least the storage 7000 series.
I think native cifs on Solaris vs. Samba on Linux is fair simply because
it''s what someone rolling out an implementation would use.
It''ll never be 100% apples-to-apples, so I''d say real-world is
preferred over hampering one system to make it *closer* to the other.
As for configurations, I probably have access to enough hardware to do most of
the benchmarking, but this time of year, being end-of-quarter, I
wouldn''t have the time to do so. That doesn''t mean there
isn''t someone else lurking who does.
--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/d800756c/attachment.html>
On Wed, Dec 3, 2008 at 4:15 PM, Joseph Zhou <jz at excelsioritsolutions.com>wrote:> haha, Tim, yes, I see the Open spirit in this reply! ;-) > > As I said, I am just exploring data. > > The Sun J4000 SPC1 and SPC2 benchmark results were nice, just lacking other > published results with the iSCSI HBA as DAS, not as a network storage device > (as 7000). Though I would attempt to say those results can be a basis for > 7000 block-performance... > > any comment? > Thanks! > z >I''d imagine you''ll see far better performance out of the 7000 with their use of flash. Only time will tell though :) --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081203/662fb08e/attachment.html>
Joseph Zhou wrote:> Thanks Ian, Tim, > Ok, let me really hit one topic instead of trying to see in general > what data are out there... > > Let''s say OpenSolaris doing Samba vs. Linux doing Samba, in CIFS > performance. > (so I can link to the Win2008 CIFS numbers and NetApp CIFS numbers > myself.) > > Is there any data to this specific point?I think what we are telling you is the only way to find the numbers you want for your configuration is to do your own tests. There are just too many variables for other people''s data to be truly relevant. One of the benefits of Open Source is you only have to pay for your time to run tests. As Tim said, there''s no point in limiting OpenSolaris to Samba. -- Ian.
Ian, Tim, again, thank you very much in answering my question. I am a bit disappointed that the whole discussion group does not have one person to stand up and say "yeah, OpenSolaris absolutely outperforms Linux and Windows, because......" But I wish, one day, we can be arguing not on a basis of belief, but on a basis of facts (referencable data). I can test all I want, the results don''t mean anything in official arguments because I am not VERITEST, and my firm is not funding my testings. With all the love for Sun Storage, and all the disappointments, please, keep this in mind. Thank you! zStorageAnalyst ----- Original Message ----- From: "Ian Collins" <ian at ianshome.com> To: "Joseph Zhou" <jz at excelsioritsolutions.com> Cc: "Tim" <tim at tcsac.net>; <zfs-discuss at opensolaris.org> Sent: Wednesday, December 03, 2008 5:43 PM Subject: Re: [zfs-discuss] OpenSolaris vs Linux> Joseph Zhou wrote: >> Thanks Ian, Tim, >> Ok, let me really hit one topic instead of trying to see in general >> what data are out there... >> >> Let''s say OpenSolaris doing Samba vs. Linux doing Samba, in CIFS >> performance. >> (so I can link to the Win2008 CIFS numbers and NetApp CIFS numbers >> myself.) >> >> Is there any data to this specific point? > > I think what we are telling you is the only way to find the numbers you > want for your configuration is to do your own tests. There are just too > many variables for other people''s data to be truly relevant. > > One of the benefits of Open Source is you only have to pay for your time > to run tests. > > As Tim said, there''s no point in limiting OpenSolaris to Samba. > > -- > Ian. >
Tim, thanks, yeah, I have highlighted Sun Storage SSD, see my blog, if you are
really interested.
http://ideasint.blogs.com/ideasinsights/2008/10/ssd-shines-new.html
note the Sun Storage comment to the blog and my reply.
Happy holidays!
z
----- Original Message -----
From: Tim
To: Joseph Zhou
Cc: Ian Collins ; zfs-discuss at opensolaris.org
Sent: Wednesday, December 03, 2008 5:22 PM
Subject: Re: [zfs-discuss] OpenSolaris vs Linux
On Wed, Dec 3, 2008 at 4:15 PM, Joseph Zhou <jz at
excelsioritsolutions.com> wrote:
haha, Tim, yes, I see the Open spirit in this reply! ;-)
As I said, I am just exploring data.
The Sun J4000 SPC1 and SPC2 benchmark results were nice, just lacking other
published results with the iSCSI HBA as DAS, not as a network storage device (as
7000). Though I would attempt to say those results can be a basis for 7000
block-performance...
any comment?
Thanks!
z
I''d imagine you''ll see far better performance out of the
7000 with their use of flash. Only time will tell though :)
--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20081206/ddc9eb08/attachment.html>
On Sat, 06 Dec 2008 22:28:36 -0500 Joseph Zhou <jz at excelsioritsolutions.com> wrote:> Ian, Tim, again, thank you very much in answering my question. > > I am a bit disappointed that the whole discussion group does not have > one person to stand up and say "yeah, OpenSolaris absolutely > outperforms Linux and Windows, because......"Why? What purpose would it serve? For some tasks Linux outperforms Windows and OpenSolaris. For some tasks Windows outperforms OpenSolaris and linux. For some tasks OpenSolaris outperforms linux and Windows.> But I wish, one day, we can be arguing not on a basis of belief, but > on a basis of facts (referencable data).So you''re discounting all the publicly available information that''s not only on sun.com, but also blogs.sun.com (see, eg, Roch''s and R.Elling''s blogs), and joyent, and many other places. Why is that? One thing that I find quite refreshing about these fora is that there is a distinct preference for hard, referencable data, and the intestinal fortitude to analyse it objectively. James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog
"James C. McPherson" <James.McPherson at Sun.COM> wrote:> On Sat, 06 Dec 2008 22:28:36 -0500 > Joseph Zhou <jz at excelsioritsolutions.com> wrote: > > > Ian, Tim, again, thank you very much in answering my question. > > > > I am a bit disappointed that the whole discussion group does not have > > one person to stand up and say "yeah, OpenSolaris absolutely > > outperforms Linux and Windows, because......" > > Why? What purpose would it serve? For some tasks Linux outperforms > Windows and OpenSolaris. For some tasks Windows outperforms OpenSolaris > and linux. For some tasks OpenSolaris outperforms linux and Windows.And in any case, the numbers do not tell useful things if you did not run the right tests. If you e.g. meter the time to unpack a tar archive using GNU tar on Linux, you cannot tell what you metered at all as the related actions in the OS kernel are not in sync with the runtime of GNU tar. An OS that feels slower may actuall be much faster just because people have subjective impressions and because one OS may have been optiomized to result in best subjective impressions only. J?rg -- EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin js at cs.tu-berlin.de (uni) schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
On Mon, 8 Dec 2008, Joerg Schilling wrote:> An OS that feels slower may actuall be much faster just because > people have subjective impressions and because one OS may have been > optiomized to result in best subjective impressions only.Microsoft Windows is surely faster than any Unix because it executes the "foreground" task (the program with the highlighted title bar) with far more priority than any other task. Microsoft Windows XP Home Edition is fastest since it maximally cranks up the priority on the foreground task given that home users can only realistically work on one thing at a time with their 17" displays. Generally people find that server oriented OSs feel "slower" because they are optimized to do more than one thing at once. They are optimized for throughput rather than instantaneous response time. Server OSs focus on providing data caching which is likely to speed up the next similar request. Regardless of many people''s wrong impressions, Solaris and OpenSolaris are a server/enterprise type OS intended to sustain heavy multi-user application loads, and not a desktop productivity OS. OpenSolaris just puts more modern Linux-like decorations on top than Solaris 10. The tiger has not lost its stripes. Linux does reasonably well for desktop productivity and does quite well at implementing the popular "LAMP" server configuration. The "LAMP" server configuration is not particularly demanding and any server limitations may be solved by simply replicating the server a few more times. I have no doubt that Linux will match Solaris performance in small LAMP style servers, and also have no doubt that Solaris will excell when faced with huge application work loads while Linux suffers. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
>Microsoft Windows is surely faster than any Unix because it executes >the "foreground" task (the program with the highlighted title bar) >with far more priority than any other task. Microsoft Windows XP Home >Edition is fastest since it maximally cranks up the priority on the >foreground task given that home users can only realistically work on >one thing at a time with their 17" displays.Solaris does the same thing. (The X server will run the "foreground" processes with a higher priority) Casper
On Mon, 8 Dec 2008, Casper.Dik at Sun.COM wrote:> > Solaris does the same thing. > > (The X server will run the "foreground" processes with a higher priority)Yes, it does, but I suspect not to the extreme degree as seen under Windows. The performance difference between the "foreground" and background processes under Windows is quite extreme. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
Hi, A couple of years ago I compared Solaris 10 and XP x64 on the same hardware (dual opteron) running the same analytical cases. Each OS had a clean install before use :- Case CPU time System Time Lapse Bits OS No Secs Secs hh:mm:ss 1 5890 2 1:38:37 64 Solaris 10 2 5578 2 1:38:31 64 Solaris 10 3 1128 1 0:19:05 64 Solaris 10 1 6536 2 1:49:19 32 Solaris 10 2 8388 2 2:20:12 32 Solaris 10 3 1311 1 0:22:11 32 Solaris 10 1 0 0 3:16:24.21 32 Win XP x64 2 0 0 3:17:27.71 32 Win XP x64 3 2220 0 0:37:00 32 Win XP x64 The tests were repeated just to make sure. Unfortunately until software is built and tested on Solaris 10, people tend to assume Windows is faster. The above tests were completed with no virus scanner installed as not to distort the results. -- This message posted from opensolaris.org
On Tue, 9 Dec 2008, russell aspinwall wrote:> The tests were repeated just to make sure. Unfortunately until > software is built and tested on Solaris 10, people tend to assume > Windows is faster. The above tests were completed with no virus > scanner installed as not to distort the results.Your data is surely out of date. Windows itself inserts "anti-virus" type checking into your application as it runs. Windows executes your application slower with each new service pack update and more and more run-time safety checks are added. If you build your application with recent Visual Studio versions, then it may run vastly slower by default. Adobe developers found that debug versions of Photoshop became completely unusable due to all the added background checking. Much of the checking can be turned off, but it may require rebuilding the SDKs with a special define. There may be specialized areas where Windows does shine, but it is clearly not in application execution times. It is possible that someone will develop a Solaris-specific virus, but that seems unlikely. Windows is paying a high price for its policies. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
Bob Friesenhahn <bfriesen at simple.dallas.tx.us> wrote:> Your data is surely out of date. Windows itself inserts "anti-virus" > type checking into your application as it runs. Windows executes your > application slower with each new service pack update and more and more > run-time safety checks are added. If you build your application with > recent Visual Studio versions, then it may run vastly slower byAre you talking about the stack overflow checking that is added by the compiler? J?rg -- EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin js at cs.tu-berlin.de (uni) schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
Russell, thanks much for the testing data! Although out of date and not a server data, not what I am looking for, the data is valuable on the thread for others that may have the need. And one day, we can have "open storage validated open performance data" as a brand.... anyway, since we are talking about OpenSolaris vs Windows now, I do have another request to the list, on NAS protocol and performance. Any comments on OpenSolaris native CIFS vs. Windows SMB 2.0? Data or theory, all welcome, thanks! See below for tech history, if not already clear. I doubt many folks still remember IBM invented SMB (CIFS) and Sun invented NFS, and NetApp did not invent NAS! zStorageAnalyst Barry Feigenbaum originally designed Server Message Block (SMB) at IBM to enable local file-access into a networked file-system. Microsoft merged the SMB protocol with the LAN Manager product and continued to add features to the protocol in every version of Windows. At around the time when Sun Microsystems announced WebNFS (an extension to the NFS file system), Microsoft launched an initiative in 1996 to rename SMB to Common Internet File System (CIFS), and added more advanced features. Microsoft submitted some partial specifications as Internet-Drafts to the IETF though these submissions have expired. Because of the importance of the SMB protocol in Microsoft Windows platform, from the open source camp, the Samba project originated with the aim of reverse engineering and providing a free implementation of a compatible SMB client and server for use with non-Microsoft operating systems. With Windows Vista, Microsoft introduced SMB 2.0. And with Windows Server 2008, SMB 2.0 again the official name for MS file service protocol, not CIFS again. And MS claims Windows SMB 2.0 bring huge performance gain. ----- Original Message ----- From: "Joerg Schilling" <Joerg.Schilling at fokus.fraunhofer.de> To: <russell.aspinwall at flomerics.co.uk>; <bfriesen at simple.dallas.tx.us> Cc: <zfs-discuss at opensolaris.org> Sent: Tuesday, December 09, 2008 2:16 PM Subject: Re: [zfs-discuss] OpenSolaris vs Linux Bob Friesenhahn <bfriesen at simple.dallas.tx.us> wrote:> Your data is surely out of date. Windows itself inserts "anti-virus" > type checking into your application as it runs. Windows executes your > application slower with each new service pack update and more and more > run-time safety checks are added. If you build your application with > recent Visual Studio versions, then it may run vastly slower byAre you talking about the stack overflow checking that is added by the compiler? J?rg -- EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin js at cs.tu-berlin.de (uni) schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily _______________________________________________ zfs-discuss mailing list zfs-discuss at opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss