Popularity of the current Solaris releases
Dave
* Solaris 10 on SPARC processors
* Solaris 10 on x86 processors
* OpenSolaris on SPARC processors
* OpenSolaris on x86 processors.
how do they compare in popularity?
I assume it fairly safe to assume that OpenSolaris on SPARC is less popular than
OpenSolaris on x86, but I'm less sure how the others rank.
I guess Solaris 10 is more popular on SPARC than x86, but maybe that is not true.
--
I respectfully request that this message is not archived by companies as
unscrupulous as 'Experts Exchange' . In case you are unaware,
'Experts Exchange' take questions posted on the web and try to find
idiots stupid enough to pay for the answers, which were posted freely
by others. They are leeches.
John D Groenveld
>There are basically 4 current Solaris operating systems?
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
>how do they compare in popularity?
My wild-assed guess is that it is heavily dependent on the
business sector and the application.
For instance, you can look at your web server's agent_log's
or its Google Analytics numbers but I suspect browsers on
stationary and mobile developer workstations are mostly x86
running Indiana/Nevada.
Obviously that doesn't really help you when you're trying
gauge how large a barrier an OpenSolaris dependency might
place on the usage of your sage-math bits inside data centers.
Netcraft's numbers, particularly the .EDU industry numbers
might be a bit more helpful, but I think you have to pay
for that level of detail.
My WAG is that OpenSolaris is not a major barrier for your
potential users, but I didn't attend Sun's HPC consortium
meeting at Super Computing this year so I'm only surmising
from my tiny little corner of the world.
Sun's HPC marketing wonks host user group meetings in Europe.
I suggest you join that community if you haven't already.
<URL:http://hpc.sun.com/>
You might find some folks there willing to donate you zones
on their various systems for your porting and benchmarking
efforts.
Happy hacking,
John
groe...@acm.org
Canuck57
> There are basically 4 current Solaris operating systems?
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
> how do they compare in popularity?
>
> I assume it fairly safe to assume that OpenSolaris on SPARC is less
> popular than OpenSolaris on x86, but I'm less sure how the others rank.
>
> I guess Solaris 10 is more popular on SPARC than x86, but maybe that is
> not true.
My guess for business is:
1) Solaris 10 on SPARC processors
2) Solaris 10 on x86 processors
3) OpenSolaris on x86 processors.
4) OpenSolaris on SPARC processors (and could be near zero)
For hobbiests and at home:
1) Solaris 10 on x86 processors
2) OpenSolaris on x86 processors.
3) Solaris 10 on SPARC processors
4) OpenSolaris on SPARC processors (and could be near zero)
In each case, the top 2 represent 98% of the cases.
The best thing about Solaris on AMD/intel is simple, you can learn
skills that port well to big iron. Even develop nice apps on x86 and
port to Sparc for the business side.
I also suspect x86 for business servers is also growing at Sun, or
fairing better than Sparc as a percentage of sales.
Richard B. Gilbert
> There are basically 4 current Solaris operating systems?
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
> how do they compare in popularity?
>
> I assume it fairly safe to assume that OpenSolaris on SPARC is less
> popular than OpenSolaris on x86, but I'm less sure how the others rank.
>
> I guess Solaris 10 is more popular on SPARC than x86, but maybe that is
> not true.
>
>
I think popularity is not terribly important. It does matter, big time,
that the O/S runs on your hardware and supports most peripherals. It
matters that the system can get your work done in a reasonable amount of
time. Whether "reasonable" means seconds, minutes, or hours depends on
both your needs and your budget.
It also matters that the O/S is reasonably free of bugs and security
vulnerabilities.
Dave
> Dave wrote:
>> There are basically 4 current Solaris operating systems?
>>
>> * Solaris 10 on SPARC processors
>> * Solaris 10 on x86 processors
>> * OpenSolaris on SPARC processors
>> * OpenSolaris on x86 processors.
>>
>> how do they compare in popularity?
>>
>> I assume it fairly safe to assume that OpenSolaris on SPARC is less
>> popular than OpenSolaris on x86, but I'm less sure how the others rank.
>>
>> I guess Solaris 10 is more popular on SPARC than x86, but maybe that
>> is not true.
>
> My guess for business is:
>
> 1) Solaris 10 on SPARC processors
My feeling is that would be top too.
> 2) Solaris 10 on x86 processors
> 3) OpenSolaris on x86 processors.
> 4) OpenSolaris on SPARC processors (and could be near zero)
My feeling that would be bottom too.
> For hobbiests and at home:
>
> 1) Solaris 10 on x86 processors
> 2) OpenSolaris on x86 processors.
I would have thought that OpenSolaris (on x86) is much more popular for
home/hobbyists than Solaris 10 on x86, simply because OpenSolaris has a better
desktop experience and better support for drivers for commodity x86 hardware.
I guess if your reason for running Solaris is to learn something that you might
be able to use for employment, Solaris 10 would be a better choice, but then I
would have thought it better to buy a SPARC in that case.
> 3) Solaris 10 on SPARC processors
> 4) OpenSolaris on SPARC processors (and could be near zero)
Yes, I get the feeling that OpenSolaris on SPARC is near zero.
> In each case, the top 2 represent 98% of the cases.
>
> The best thing about Solaris on AMD/intel is simple, you can learn
> skills that port well to big iron. Even develop nice apps on x86 and
> port to Sparc for the business side.
> I also suspect x86 for business servers is also growing at Sun, or
> fairing better than Sparc as a percentage of sales.
Probably true, as performance per � is better. It's a shame for Sun, as I'm led
to believe the profit on x86 is small compared to SPARC. That said, judging by
the amount of money Sun wanted for RAM and disks for my Ultra 27 (Intel Xeon),
was it any surprise I bought it with 500 GB disk and 2 GB RAM? (I did consider
buying with no disk or RAM, but thought that would be more difficult if a
warranty issue arose).
I bought the other 10 GB from Crucial, and the enterprise grade disks elsewhere,
for a hell of a lot less than Sun wanted for them.
Dave
My issue is one of developing open-source software (Sage), to run on hardware
others have. Some will use it as hobbyists, others as departmental servers,
supporting 100's of students.
My first priority was SPARC on Solaris 10, as that is what I run, and that is
what the T5240 donated by Sun runs. Next will be OpenSolaris on x86, simply as I
have an Ultra 27 myself.
Believe it or not, I'm doing a bit on HP-UX in parallel, mainly as different
operating systems often show bugs which another platform does not. I'm sure its
fastest to develop software on linux, and only worry about it building on Linux.
I'm also pretty convinced you gain in quality by addressing portability issues,
and testing on multiple platforms.
Dave
Richard B. Gilbert
Just don't expect Sun to service third party hardware. You'll need to
keep your own spares and be prepared to replace disks or memory DIMMs at
inconvenient times!
It's not a huge burden. Just make sure that you have a stock of spare
parts adequate for your need. And make sure you know how to open the
case and replace a disk.
Canuck57
I will give Sun this, their memory prices are better than IBM or HP, but
then their sales are in the tank for UNIX iron.
But still, the pricing on Sparc hurts. But at least Sun's stuff works
with COTs parts. I too have put memory and disk from COT suppliers. You
are at waranty risk but saving $20,000 on a base $20,000 machine is
often the benifit. It is that or some sucky MS-Windows boxen from hell.
Canuck57
Agreed. But one hold back for Solaris on x86 is the lack of SATA
ICHR7/8/9/10 drivers in modes other than IDE emulation, which many BIOS
no longer supports the IDE downgrade part.
I have 3 perfectly good common mobo based systems at home, none will run
Solaris nor OpenSolaris outside a VM. And the chipsets are amongst the
most common in the desktop quad processor arena there is.
A severe handycap for Solaris adoption. Part of why I am increasingly
recommending Linux, as not every app is going to fork $20k+++$$$$ and up
for a Sparc with decent memory/storage. Solaris seems to be taking the
path to a niche OS. Unfortunate, as I think it is the best. Just being
wasted like OS/2. OS/2 dies for the same reasons, crappy driver support.
Canuck57
Good technique too. Cheap, fast and easy way to find problematic areas
of code is to look at the reasons why it will not cross compile
unaltered. Not sure I would worry about Itanic though. Bad idea from
HP to drop PA/RISC. Killed them in the real server market.
Paul Floyd
> often the benifit. It is that or some sucky MS-Windows boxen from hell.
Is that a typo for dell? Or are they the same thing anyway?
A bientot
Paul
--
Paul Floyd http://paulf.free.fr
Bill Waddington
wrote:
>Richard B. Gilbert wrote:
>> Dave wrote:
>Agreed. But one hold back for Solaris on x86 is the lack of SATA
>ICHR7/8/9/10 drivers in modes other than IDE emulation, which many BIOS
>no longer supports the IDE downgrade part.
>
>I have 3 perfectly good common mobo based systems at home, none will run
>Solaris nor OpenSolaris outside a VM. And the chipsets are amongst the
>most common in the desktop quad processor arena there is.
Pardon my ignorance, but is there something different about the ICH10R
chipset? Or maybe I'm misreading this.
I've got an EVGA i7 x58 mobo happily booting S10 U7 and SNV 113. The
SATA mode in BIOS is straight AHCI for the hard drives and opticals.
Not IDE emulation.
Bill
--
William D Waddington
william.w...@beezmo.com
"Even bugs...are unexpected signposts on
the long road of creativity..." - Ken Burtch
Thomas Tornblom
> Agreed. But one hold back for Solaris on x86 is the lack of SATA
> ICHR7/8/9/10 drivers in modes other than IDE emulation, which many
> BIOS no longer supports the IDE downgrade part.
>
> I have 3 perfectly good common mobo based systems at home, none will
> run Solaris nor OpenSolaris outside a VM. And the chipsets are
> amongst the most common in the desktop quad processor arena there is.
>
I don't know what system boards you're using, but my cheap intel
DG33TL happily uses native SATA ion its ICH9R controller with both
Solaris Express and Opensolaris:
From scanpci:
pci bus 0x0000 cardnum 0x1f function 0x02: vendor 0x8086 device 0x2922
Intel Corporation 82801IR/IO/IH (ICH9R/DO/DH) 6 port SATA AHCI Controller
Have you enabled SATA mode in the BIOS? The default appears to be IDE
emulation on many systems.
I have friends happily using SATA on other boards with Solaris, so
your knowledge appears to be dated.
Have you run SDDTool?
"Use this tool to determine whether the devices that are detected in
your system are supported by the drivers that exist in the Solaris OS
for the x86/x64 platform."
Dave
In my case, I've not paid for a support contract - but are covered by things
like the Sale of Goods Act in the UK.
I've kept the original 500 GB disk, but I've mirrored it with an identical part
number from Hitachi. I'm still using the original 2 GB DIMM.
Dave
> Dave wrote:
>> I bought the other 10 GB from Crucial, and the enterprise grade disks
>> elsewhere, for a hell of a lot less than Sun wanted for them.
>
> I will give Sun this, their memory prices are better than IBM or HP, but
> then their sales are in the tank for UNIX iron.
>
> But still, the pricing on Sparc hurts. But at least Sun's stuff works
> with COTs parts. I too have put memory and disk from COT suppliers. You
> are at waranty risk but saving $20,000 on a base $20,000 machine is
> often the benifit. It is that or some sucky MS-Windows boxen from hell.
Does anything beat what HP wanted for RAM to upgrade my HP 1320nw Postscript
printer? Back in 2006, then wanted �680 to add 128 MB, which was far more than
I'd paid for printer. It worked out at �5440 per GB or $9646 / GB (based on
exchange rate in 2006). I ended up pay �29.99 for 128 BM, which made HP RAM more
than 22x the cost of Crucial.
This was my post on the subject back in 2006.
Sun wanted a bit over twice Crucial for the RAM for the Ultra 27.
Thomas Tornblom
> Canuck57 wrote:
>> Dave wrote:
>
>>> I bought the other 10 GB from Crucial, and the enterprise grade
>>> disks elsewhere, for a hell of a lot less than Sun wanted for them.
>> I will give Sun this, their memory prices are better than IBM or HP,
>> but then their sales are in the tank for UNIX iron.
>> But still, the pricing on Sparc hurts. But at least Sun's stuff
>> works with COTs parts. I too have put memory and disk from COT
>> suppliers. You are at waranty risk but saving $20,000 on a base
>> $20,000 machine is often the benifit. It is that or some sucky
>> MS-Windows boxen from hell.
>
> Does anything beat what HP wanted for RAM to upgrade my HP 1320nw
> Postscript printer? Back in 2006, then wanted �680 to add 128 MB,
> which was far more than I'd paid for printer. It worked out at �5440
> per GB or $9646 / GB (based on exchange rate in 2006). I ended up pay
> �29.99 for 128 BM, which made HP RAM more than 22x the cost of Crucial.
>
I just added 128M to the 96M I had in my old HP Color Laserjet 4500 and
paid a whopping $21, plus $5.50 for international shipping :-)
Thomas
Andrew Gabriel
Canuck57 <Canu...@nospam.com> writes:
> Agreed. But one hold back for Solaris on x86 is the lack of SATA
> ICHR7/8/9/10 drivers in modes other than IDE emulation, which many BIOS
> no longer supports the IDE downgrade part.
Solaris has driven these natively for some time with ahci(7D).
They're used in a number of Sun systems, and probably half
the newer systems OpenSolaris is routinely installed on.
> I have 3 perfectly good common mobo based systems at home, none will run
> Solaris nor OpenSolaris outside a VM. And the chipsets are amongst the
> most common in the desktop quad processor arena there is.
Sounds like you have poor motherboards and/or ancient BIOS.
Probably worth trying a BIOS update, if the board manufacturer
makes them available (sadly some cheaper ones don't after they
stop manufacturing the board, and you're stuck forever more
with whatever bugs the BIOS has at that point).
If you said what motherboard models you're using, someone
else might say if they have them working OK.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
Richard L. Hamilton
Dave <f...@coo.com> writes:
> There are basically 4 current Solaris operating systems?
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
> how do they compare in popularity?
>
> I assume it fairly safe to assume that OpenSolaris on SPARC is less popular than
> OpenSolaris on x86, but I'm less sure how the others rank.
>
> I guess Solaris 10 is more popular on SPARC than x86, but maybe that is not true.
OpenSolaris (the distro) is a new development on SPARC, and support for
older graphics hardware is poor-to-nonexistent, so you have to figure that
combination would presently be quite rare, esp. on desktops.
Richard L. Hamilton
Well...long ago, a lot of development was done on Suns. Porting from them to
something else was usually pretty easy. Then Linux came along, with the attitude
problem that since it's all open source, everybody else should be too, and then
it would all just work. So rather than reading standards (XPG, POSIX, SUS, etc),
a lot of folks doing application development on Linux assume that Linux _is_
a standard rather than just an OS implementation, and don't give a hoot about
portability.
That sort of thing, whether by attitude or carelessness, is greatly
mitigated by actually porting the code to something else. Nevertheless, if
sufficient care is taken to use standards-defined interfaces as much as
possible and isolate any code that needs anything more, one can end up in
the same place (arguably with cleaner code, too), even prior to porting to
additional platforms. But systemic understanding and discipline aren't as
popular as they used to be (if they ever were...). :-/
solx
> There are basically 4 current Solaris operating systems?
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
> how do they compare in popularity?
>
> I assume it fairly safe to assume that OpenSolaris on SPARC is less
> popular than OpenSolaris on x86, but I'm less sure how the others rank.
>
> I guess Solaris 10 is more popular on SPARC than x86, but maybe that is
> not true.
>
>
Hi,
Having worked for a software house, Sun has done themselves no favours
by dropping SPARC workstations. The Sun workstations are all AMD or
Intel based, while servers are SPARC, AMD or Intel.
I worked with developers who wanted SPARC workstations but were given
AMD64/x86 based workstations running Windows. Hopefully with Solaris 11
Sun will release a SPARC workstation, other wise I would expect to see
Solaris/OpenSolaris to run primarily on AMD/Intel hardware. The SPARC
chip appears to be going the way of the old Motorola 68000 series chips
Sun used to use before introducing the SPARC, the SPARC will give way to
the AMD/Intel chips. They tried it in the late 1980's with the old 80386
but the AMD/Intel chips are considerably more powerful now.
Martha Starkey
Andrew Gabriel
>
> * Solaris 10 on SPARC processors
> * Solaris 10 on x86 processors
> * OpenSolaris on SPARC processors
> * OpenSolaris on x86 processors.
>
> how do they compare in popularity?
Depends very heavily on the type of customer base you're considering.
> I assume it fairly safe to assume that OpenSolaris on SPARC is less popular than
> OpenSolaris on x86, but I'm less sure how the others rank.
>
> I guess Solaris 10 is more popular on SPARC than x86, but maybe that is not true.
You also need to define what you mean by popularity.
Andrew Gabriel
solx <nos...@example.net> writes:
> Hi,
>
> Having worked for a software house, Sun has done themselves no favours
> by dropping SPARC workstations. The Sun workstations are all AMD or
> Intel based, while servers are SPARC, AMD or Intel.
> I worked with developers who wanted SPARC workstations but were given
What did they want them for?
If you want to develop for sparc servers, you need to do that on
a sparc server. They are now so different from workstations, that
if you developed on a workstation, your app would probably run
very badly on a modern server.
> AMD64/x86 based workstations running Windows. Hopefully with Solaris 11
Stick Solaris x86 on it and use it to front-up your sparc server
for same look and feel, or even just as an X terminal (or use a
SunRay). There's no shortage of ways to achieve the same thing,
but developing on a sparc workstation won't any longer give you
any feel for how a modern server works - they've moved on too
far from simply being big versions of a workstation.
Drazen Kacar
> They are now so different from workstations, that if you developed on a
> workstation, your app would probably run very badly on a modern server.
Most apps run very badly. Then you buy more hardware. :-)
--
.-. .-. Yes, I am an agent of Satan, but my duties are largely
(_ \ / _) ceremonial.
|
| da...@fly.srk.fer.hr
Dave
Are you referring in particular to the CoolThreads chips? They certainly are a
totally different animal to the SPARCs in any workstations I've ever used. They
clearly are capable of high performance if your code can exploit their
parallelism, or you have a lot of processes, but otherwise they are pretty damm
slow. I've just about given up trying to develop on a T5240 due to its speed.
Andrew Gabriel
T series and M series - they're very different beasts, but again both
very different from a workstation. For a T5240, you're going to need
somewhere up around 128 runnable threads to get the most from it,
and ideally many more in practice. That's great for throughput
computing such as web serving, and apps which have been designed to
scale to very many threads such as Oracle*, but if you write an app
on a workstation and just move it across without understanding the
differences, you might find you can only use about 1% of the server's
capability. For M-series, you also need a good number of runnable
threads, 16 or more, and you have that combined with cores that each
give you something like 3 times the performance of something like a
V890 core at the same clockspeed. These systems just don't look
anything like big workstations anymore - they've well outgrown that.
I suppose the nearest would be an M3000 with a SunRay.
* http://www.oracle.com/us/corporate/press/036544
yourmommycalled
> In article <4b15b...@212.67.96.135>,> >> In article <Toqdne9RdMmsbo7WnZ2dnUVZ8kqdn...@pipex.net>,
>
> --
> Andrew Gabriel
> [email address is not usable -- followup in the newsgroup]
For some us 16 cores running at 3x the speed of a UltraSparc IV+ would
be very nice. Think WRF-DA with 53 vertical layers and 300m horizontal
resolution
Dave
>> Are you referring in particular to the CoolThreads chips? They certainly are a
>> totally different animal to the SPARCs in any workstations I've ever used. They
>> clearly are capable of high performance if your code can exploit their
>> parallelism, or you have a lot of processes, but otherwise they are pretty damm
>> slow. I've just about given up trying to develop on a T5240 due to its speed.
>
> T series and M series - they're very different beasts, but again both
> very different from a workstation. For a T5240, you're going to need
> somewhere up around 128 runnable threads to get the most from it,
> and ideally many more in practice. That's great for throughput
> computing such as web serving, and apps which have been designed to
> scale to very many threads such as Oracle*, but if you write an app
> on a workstation and just move it across without understanding the
> differences, you might find you can only use about 1% of the server's
> capability. For M-series, you also need a good number of runnable
> threads, 16 or more, and you have that combined with cores that each
> give you something like 3 times the performance of something like a
> V890 core at the same clockspeed. These systems just don't look
> anything like big workstations anymore - they've well outgrown that.
> I suppose the nearest would be an M3000 with a SunRay.
>
> * http://www.oracle.com/us/corporate/press/036544
>
I fully accept that programs like Oracle and Apache make good use of machines
with CPUs with a low clock speed, but many threads.
Unfortunately, computers based on such processors can present problems for many
reasons.
1) There may be simply no way to exploit the parallelism when trying to solve a
specific problem - some things simply can't be done in parallel.
2) There may be no known algorithm to exploit the parallelism, though one exists.
3) A parallel algorithm may be known, but is significantly more difficult to
implement than a serial one.
4) A parallel implementation of a program is possible, but the parts that can't
be done in parallel will dominate the CPU time, so the speedup from making use
of lots of threads is quite small.
5) Multi-threaded programs are significantly more difficult to debug than serial
ones, as failures tend not to be so reproducible.
6) You may be using software written by someone else, which could exploit
parallel processors, but which does not. Even if you have access to the source
code, you will probably not want to devote the time to convert their program to
something which exploits multiple processors.
There are probably other reasons too, but those 6 come quickly to my mind.
However, I am well aware, that making a CPU run at 1 GHz with 64 threads, is
easily practical today, whereas making a CPU run at 64 GHz is not possible today
and probably never will be.
Dave
David Combs
...
>
>I bought the other 10 GB from Crucial, and the enterprise grade disks elsewhere,
For your site (or thought), what brands of disks qualify as "enterprise" grade?
And are those SCSI (iscsi?) or other?
Also, is the general understanding that the underlying DISK-HARDWARE for
scsi disks is more robust?
Thanks!
David
David Combs
Who did you buy it from, HP itself?
David
David Combs
>fastest to develop software on linux,
Why? What is that you experience with linux that makes it better for
software development than solaris?
Thanks,
David
David Combs
Wow, is that statement a surprise!
(1) A Blade-2500(red): workstation or server?
(2) Please, elaborate a bit (maybe a lot!) on
the difference between a "sparc workstation" (were Sun to
resume making them) and a "server", as far as developing
software via the workstation to eventually run on
the server.
(3) I always thought solaris was solaris was solaris, be it
workstation or server. Where was I wrong?
THANKS!
David (and many others equally surprised!)
Tim Bradshaw
> 2) Please, elaborate a bit (maybe a lot!) on
> the difference between a "sparc workstation" (were Sun to
> resume making them) and a "server", as far as developing
> software via the workstation to eventually run on
> the server.
A T-series box, for instance, has oodles of rather slow "cpus", whereas
a workstation has only two or 4 or something, and they are probably
individually faster.
David Combs
But how would that make them (sparc workstations) useless
for writing server programs?
David
Ian Collins
Why?
--
Ian Collins
Tim Bradshaw
> But how would that make them (sparc workstations) useless
> for writing server programs?
It makes it very hard to do any kind of performance tests. The likely
result of that is that when the application gets deployed on a T-series
machine, the performance is terrible because there's not enough
parallelism, or equivalently there are single-threaded bottlenecks all
over it. Very smart people might be able to deal with this at the
design stage (ie "I know the performance of this application will be
terrible on the machine I write it on, but I'm clever enough to be able
to predict that it will be OK on the target platform"), but I'm not
sure I've ever met anyone that clever.