Slashdot Mirror

Check it out, more volume, size of a walkman and cheaper as well. compaq unit

Compaq has lots of good (free) information in their ActiveAnswers section. Specifically referring to Linux & Apache

Basically you seperate the web servers from the data from the application servers (cgi). Their testing revealed that it scales very well (adding a server under DISA typically created more benefit than if each server was self-contained). This will probably require more hardware initially but will allow you to add power where needed.

As for bandwidth, more is always good, but you should be able to use tools like mrtg to help you determine if you've got the optimal + headroom or not. If you can co-lo you might reduce some of the initial expenses (no need for a data room, getting the telco to bring the pipes in, power, etc) but you'll probably lose some flexibility.

It's worse, Compaq is calling their machine the "iPaq." Oy. Check out http://www.compaq.com/produ cts/internetdevices/index.html for details. I love the part about "Machines configured with Windows 2000 will ship when the operating system is available."

Ok, here goes:

1. The Smart Array driver is *not* a binary only driver

2. Yes, there is a kernel patch. It's in 2.2.12 or 13, and 2.3.x. You can also get it from ftp://ftp.compaq.com/pub/products/ drivers/linux/.

3. No. The driver was written at Compaq (by me). It is now maintained by another employee at Compaq.

4. Since I don't normally work on storage products, I don't know the answer to this. Hopefully the answer will be yes in the near future.

5. Yes. The Linux SCSI code has support for hot-swapping drives (if you're SCSI subsystem supports it, electrically). For example, if I plug a new drive into my Proliant 5500 in slot 1, I can bring the drive online with

echo "scsi add-single-device 0 0 1 0" >/proc/scsi/scsi

The order of those numbers parameters are host, channel, id, lun. Similarly, you can take a disk offline with "scsi remove-single-device params".

As for the elimination of Keyboard, video and mouse and replacing them with serial port access to the BIOS, I suggest you check out a feature known as Integrated Remote Console (most Compaq servers have this feature). A serial port is "stolen" from the machine and is used to remote video and keyboard. See http:/ /www.compaq.com/support/techpubs/user_reference_gu ides/281862-002.html for more information.

I hope this helps,
--Chris

Ok, here goes:

1. The Smart Array driver is *not* a binary only driver

2. Yes, there is a kernel patch. It's in 2.2.12 or 13, and 2.3.x. You can also get it from ftp://ftp.compaq.com/pub/products/ drivers/linux/.

3. No. The driver was written at Compaq (by me). It is now maintained by another employee at Compaq.

4. Since I don't normally work on storage products, I don't know the answer to this. Hopefully the answer will be yes in the near future.

5. Yes. The Linux SCSI code has support for hot-swapping drives (if you're SCSI subsystem supports it, electrically). For example, if I plug a new drive into my Proliant 5500 in slot 1, I can bring the drive online with

echo "scsi add-single-device 0 0 1 0" >/proc/scsi/scsi

The order of those numbers parameters are host, channel, id, lun. Similarly, you can take a disk offline with "scsi remove-single-device params".

As for the elimination of Keyboard, video and mouse and replacing them with serial port access to the BIOS, I suggest you check out a feature known as Integrated Remote Console (most Compaq servers have this feature). A serial port is "stolen" from the machine and is used to remote video and keyboard. See http:/ /www.compaq.com/support/techpubs/user_reference_gu ides/281862-002.html for more information.

I hope this helps,
--Chris

http://www.testdrive.compaq.com/tru64/index.shtml

---

The Register has the details.

Basically the problem is that Motorola is shipping a buggy product. This is very similar to problems we see all the time in software (business pressure to release something before it's really ready). Chips are never perfect when they're first fabricated. They always go through several rounds of: 1) fabricate, 2) find and fix bugs in the parts.

Of course, it's not clear exactly who screwed up here. It could be that Apple announced the G4 machines before Motorola was ready. Or maybe Motorola management decided to declare the G4 ready before the hardware engineers were ready to sign off on it. Or maybe this is just a bug they didn't find until they were already shipping parts to Apple.

The Register has the details.

Basically the problem is that Motorola is shipping a buggy product. This is very similar to problems we see all the time in software (business pressure to release something before it's really ready). Chips are never perfect when they're first fabricated. They always go through several rounds of: 1) fabricate, 2) find and fix bugs in the parts.

Of course, it's not clear exactly who screwed up here. It could be that Apple announced the G4 machines before Motorola was ready. Or maybe Motorola management decided to declare the G4 ready before the hardware engineers were ready to sign off on it. Or maybe this is just a bug they didn't find until they were already shipping parts to Apple.

It's ironic to hear an opinion like this on Slashdot. If the application is Open Source or Java, it runs on OpenVMS just fine.

There is often some porting effort required with Open Sources onto OpenVMS, but it's not bad and it's getting better all the time. DEC/Compaq have greatly improved the C compiler and libraries such that most Unix C sources just compile and run. Configuration scripts in sh or bash can be a problem, but someone has ported bash to run on OpenVMS now, so maybe that will be less of a problem in the future.

Perl builds out-of-the-box (latest.tar.gz) on OpenVMS, so there's the ever-increasing catalog of Perl applications available.

Apache was recently ported. There are also a number of excellent 'native' web servers to choose from on OpenVMS. The most popular is the OSU (Ohio State University) HTTPD. It's been running multi-threaded (Posix Threads) since 1994. At one time, it was probably the most popular multi-threaded web server in the world, although that honor probably goes to multi-threaded Apache or IIS now.

There's a Python port that I believe is pretty up-to-date, but I don't know much about this.

Compaq has the first (and only today?) 64-bit implementations of Java, for Tru64 Unix and OpenVMS. See here for information. So, the Java catalog of apps will be available with world-beating Alpha performance.

There's a team forming now to port the server part of Star Office to OpenVMS. So, OpenVMS may be a fine platform for Office apps soon.

Oracle runs on OpenVMS. So, if you need that world-beating clustering for a DB Server, you can run Oracle.

I apologize to Slashdot readers for filling up this discussion with what seems like OpenVMS marketing. Although I do work for Compaq, I'm far from holding any marketing position. I'm just a happy user and consultant on OpenVMS systems.

I suspect that this 'Anonymous Coward' is someone who works for one of the Unix vendors that seem to want to spread FUD about OpenVMS in good Holloween Document style.

Read the Holloween Documents. One of Microsoft's concerns is that they can't attack Linux with FUD regarding the future of it as a platform, as they can any other vendor's products. Standard FUD from Microsoft is to make people believe that only MS products will survive in the long term and have good application support. I've seen a number of examples of people who work for a certain Unix vendor who have been trying to spread the same kind of FUD against OpenVMS lately. This particular FUD is just not true and in the Open Source or thin-client world of the future, it's largely irrelevant anyway.

Disclaimer: Yes, I work for Compaq. No, I don't speak for Compaq.

This is the newest encarnation of the world-beating clustering technology that everybody is trying to copy. Shared everything, multiple OS instances in the same box, dynamically reassignable memory/CPU between instances, seemless clustering. Read the Galaxy overview here. VMS has been doing clustering since 1985 better than any Unix does it today and it's improved a lot since then. Yet somehow, OpenVMS gets the rap as being outmoded.

Disclosure statement: Yes, I am a Compaq employee. No, I'm not speaking for Compaq.

SUN sells workstations that will beat the G4 handily except for Altivec stuff, which probably is faster. But they also have the Enterprise 10000 can be equipped with up to 64 processors, which means it will leave a G4 in the dust.

SGI sells Onyx2 InfiniteReality2, which will beat probably anything else on heavy-duty visualisation stuff, and can be equipped with up to 128 processors.

HP makes the J-5000 workstation, which will also beat a G4 on most tasks, as well as big-ass servers with up to 128 processors.

IBM makes RS/6000 workstations and servers, which can scale up to 128 processors.

Compaq sells XP1000 workstations with a 667MHz Alpha 21264 processor, which will beat the G4 on anything that can't make very good use of Altivec, and there are places that sell dual 667MHz 21264 workstations. Compaq also has the AlphaServer GS line, which can take up to 14 21264's, probably beating the G4 on anything.

Furthermore, the Athlon probably beats the G4 on stuff that doesn't parallellise well, and an 8-way Xeon should be faster for most, if not all, things.

Unfortunately all the systems here, except the Athlon, are far, far more expensive than a G4. But you can get faster systems if you're willing to pay the price. Oh, and all of those run some Unix variant, as well as Windows NT for Alpha and Athlon/PIII.

Also, when it comes to the speed of the G4, it all depends on how useful Altivec is for your app. If it isn't useful, the G4 isn't that impressive. If it is, the G4 should be very good value for money, if Altivec is anywhere near as good as the hype claims it is.

haven't Compaq opened the source of their True64 compiler a while ago

I haven't heard anything about that; do you have a reference?

Have it been ported under Linux since then?

That appears to be in progress; this page says:

We've added Compaq's beta FORTRAN for Alpha Linux to this website and it's getting rave reviews from test drivers! Register for the test drive today and check it out! COMING SOON: SuSe Linux on Alpha and Proliant and Compaq's beta C Compiler--optimized for Alpha Linux

A few answers: 1. Quite expensive. Ofcourse it depends on the model of the chip. The 21064 could not be very expensive today. The 21164 would probably cost like two PentiumIII:s at the same frequency. The powerhouse 21264 is very expensive, to say the least.
2. The only way to meassure performance is to run your application on both chips. You could also run a benchmark and hope that the benchmark is representative of your application. On the SPEC95 benchmark suite the Alpha 21264 at 667MHz scores 37.5 on the integer part and 65.5 on the floating point part. An Athlon at 600MHZ scores 27.2 in the int part and 21.6 on the fp part. This means that it is somewhere like 40% to 3 times as fast as an athlon on similar frequencies.
3. Yes, but I think it would be hard to find a mainboard for a 21264 and chip and the memory in parts. You would probably have to buy a complete system from someone like Compaq.
4. Depends on what you mean by optimized. There are linux distros that run on these, RedHad for example. But I would not call them very optimized, as gcc produces quite inefficient code for the alpha chips. The compiler for Tru64Unix (formerly known as OSF/1) is much better.
5. Why no point 5???
6. You can get them in multiprocessor configurations up to 32 processors.
7. Industry leading performance in its class.
8. Buy it from Compaq, check:
The XP1000 is quite nice...

I want to be able to boot Linux on my aero. I wonder how much work will have to be done to get linux to boot on it?

Thats really cool though, glad to see it being done.

OBLIGATORY BEOWULF COMMENT: Lets build a beowulf of these little hand held devices and take over the world!!!

One that stands out in my mind is 5,724,226: Housing Access Door Construction for a Portable Computer Docking Station. I didn't think emachines made either portables or docking stations.

The other patents cover a variety of generic PC stuff, including one that looks a lot like the infamous XOR cursor patent. IMHO, Compaq is casting a pretty wide net here, throwing any of its patents marginally related to generic PC hardware at emachines. They wouldn't dare do this to a bigger player without expecting countersuits for violating that company's patent portfolio. I guess they think that emachines doesn't have any patent portfolio of their own.

Prediction: this will get settled out of court for "an undisclosed sum." It's entirely FUD.

As anyone in the hardware business knows, there is a huge gulf between what Intel's just announced (preparing the design to be fabricated into physical chips for the first time, traditionally called "tape-out" becasue the design data used to be shipped to the people making the masks on spools of tape) and actually being ready to ship product to customers (i.e. OEMs).

Keep in mind that this means they don't have a sungle physical Merced chip yet, which means they've only been simulated, never really tested. Although I don't know anything about Intel's internal simulation tools and methodologies, I'd be surprised if they've even been able to boot an operating system on their simulated design. I'd be amazed if they've simulated significant real-world applications. In other words, at this point Merced has had very little real-world testing. That's what the next step ("first silicon") is all about: testing out whether the design actually works in practice (and systems) rather than just in theory (under pre-silicon simulation).

That's not to say that Intel hasn't expended a huge effort already in working to eliminate as many bugs as possible before taping out (in fact, I'm sure they have), but there are always more bugs when the silicon arrives. In fact, many of the problems encountered in first silicon are effects which simulation can't (or at least doesn't) effectively capture (race conditions that earlier analysis missed, unanticipated electrical and electromagnetic effects, yield problems, etc.). These problems are worse when you're talking about a new from-scratch design. There are also bound to be other problems when you're talking about a totally new processor architecture (such as IA-64).

Usually the process between tape-out and shipping to clients goes like this:

1. Tape out
2. Wait for the chips to come back from the fab
3. Attempt to run some real software (OSs, applications, etc.) on the silicon you have
4. Discover bugs in the design
5. Update the design to fix the bugs
6. Go back to step 1 and repeat until it actually works

To me, it seems that the estimate of one quarter between tape-out and shipping samples to OEMs seems extremely optimistic. Of course, I suppose Intel could be planning on using the OEMs for debugging their first or second pass silicon (before they've made it really robust), but somehow I doubt that. I would expect that it might be more like two quarters before anybody outside Intel actually sees a physical Merced chip.

But then again, hey, what do I know.

As anyone in the hardware business knows, there is a huge gulf between what Intel's just announced (preparing the design to be fabricated into physical chips for the first time, traditionally called "tape-out" becasue the design data used to be shipped to the people making the masks on spools of tape) and actually being ready to ship product to customers (i.e. OEMs).

Keep in mind that this means they don't have a sungle physical Merced chip yet, which means they've only been simulated, never really tested. Although I don't know anything about Intel's internal simulation tools and methodologies, I'd be surprised if they've even been able to boot an operating system on their simulated design. I'd be amazed if they've simulated significant real-world applications. In other words, at this point Merced has had very little real-world testing. That's what the next step ("first silicon") is all about: testing out whether the design actually works in practice (and systems) rather than just in theory (under pre-silicon simulation).

That's not to say that Intel hasn't expended a huge effort already in working to eliminate as many bugs as possible before taping out (in fact, I'm sure they have), but there are always more bugs when the silicon arrives. In fact, many of the problems encountered in first silicon are effects which simulation can't (or at least doesn't) effectively capture (race conditions that earlier analysis missed, unanticipated electrical and electromagnetic effects, yield problems, etc.). These problems are worse when you're talking about a new from-scratch design. There are also bound to be other problems when you're talking about a totally new processor architecture (such as IA-64).

Usually the process between tape-out and shipping to clients goes like this:

1. Tape out
2. Wait for the chips to come back from the fab
3. Attempt to run some real software (OSs, applications, etc.) on the silicon you have
4. Discover bugs in the design
5. Update the design to fix the bugs
6. Go back to step 1 and repeat until it actually works

To me, it seems that the estimate of one quarter between tape-out and shipping samples to OEMs seems extremely optimistic. Of course, I suppose Intel could be planning on using the OEMs for debugging their first or second pass silicon (before they've made it really robust), but somehow I doubt that. I would expect that it might be more like two quarters before anybody outside Intel actually sees a physical Merced chip.

But then again, hey, what do I know.

So the fact that they're going to be an exhibitor at the Atlanta Linux Showcase means that they don't support Linux?

What about the Active Answers for ISPs running Linux on Compaq hardware?

Or the guide they published to help install Linux and set up the web server?

Or the Alpha Linux support?

It looks to me like Compaq does support Linux. I don't think that they're preinstalling it right now, but they are supporting it, and are even setting up call in tech support..

So the fact that they're going to be an exhibitor at the Atlanta Linux Showcase means that they don't support Linux?

What about the Active Answers for ISPs running Linux on Compaq hardware?

Or the guide they published to help install Linux and set up the web server?

Or the Alpha Linux support?

It looks to me like Compaq does support Linux. I don't think that they're preinstalling it right now, but they are supporting it, and are even setting up call in tech support..

You need to download the SoftPaq for your machine. Then run it, and it will make four floppies for you (you probably ought to have four good, formatted floppies ready to go, because if one step fails, you get to do the whole thing over again). When you get the disks done, boot from the first one.

Now here is where my memory fails me slightly. At some point, you'll have to do an inventory-ish thing. The choice should be obvious, but if not, see Compaq's site for a FAQ (they have it somewhere, but I don't ahve the URL bookmarked). The deal will run, and you'll throw another of the disks in. When it's finished, you'll have what amounts to a complete, editable "snapshot" of the system. From this snapshot, you can change whatever setting you want. You just select the hardware and choose the resources you want. You also get to see a list of what's free, so it's pretty easy.

I have an old Compaq Deskpro PPro 150 that I figured would make a good gateway machine. The only problem is that I needed to put in an additional NIC (incidentally, Bay/Netgear makes a 10/100 card called the FA310TX which has the Dec Tulip chipset; it's a great card, low CPU utilization, lot's of status indicators, and only costs about $25.00). The machine already has an onboard AMD ethernet interface, but both cards wanted IRQ 5. So I ran the SoftPaq, told it where to stick its IRQ, and everything is happy behind my gateway/firewall now. If I want to add more hardware, I just run the disks again.

I agree that the floppies are a pain, and aren't as handy as a plain ROM BIOS in some cases (my battery goes dead and I'm covered). But that doesn't mean that the old Compaq sitting in the corner is useless.

Now, booting from a floppy I don't know about. I've never bothered with that before. I'm sure it's doable, but I don't recall the SoftPaq's screen menus/features that well. It's been a while.

-B

Doesn't ecology say that nature abhors a lack of diversity?

--Just coverin' my tush as an AC.

Visit Compaq's web site at http://www.compaq.com/ and make use of the form on their " Contact Us" page.

Be sure to make the following point to them in the process: Virtually nowhere do they advertise or promote the Alpha line. No wonder it's flagging in the market! I see ads everywhere for Compaq Intel boxes but no mention of their own Alpha processor. If they would only add Alpha systems to their current ad campaigns there would be a 100% increase in ad space for Alpha.

D. Keith Higgs
CWRU. Kelvin Smith Library

Visit Compaq's web site at http://www.compaq.com/ and make use of the form on their " Contact Us" page.

Be sure to make the following point to them in the process: Virtually nowhere do they advertise or promote the Alpha line. No wonder it's flagging in the market! I see ads everywhere for Compaq Intel boxes but no mention of their own Alpha processor. If they would only add Alpha systems to their current ad campaigns there would be a 100% increase in ad space for Alpha.

D. Keith Higgs
CWRU. Kelvin Smith Library

A few important points:

• This is not the group which designs the Alpha microprocessor. We're in Shrwesbury, MA, and we are not being cut.
• This is not the group which designs the Alpha systems. They are also based in MA and are also not being cut.
• This is the group which assembles the Alpha systems, after they have been designed and after the processors have been designed and fabricated.
• Compaq still has more people designing and building systems than they need, as there's overlap between the Digital groups and Compaq groups. This is just elimination of some of that redundancy.

In other words, this is nothing to worry about. Alpha is not going anywhere.

Kenneth C. Schalk ( kenneth.schalk@compaq.com)
Alpha Development Group, CAD & Test Technologies

I recently stumbled upon a tech newsletter put out by an engineer at Compaq. It is called The Rapidly Changing Face of Computing. Enjoy.

Compaq may be joining the project, but they aren't dumping their existing product.

In fact, they just announced today that Tru64 UNIX runs on simulated Merced chips. The press release hasn't hit their web page yet, though.

$450 for an 8M, TFT display, audio in/out jack, 10-hour rechargeable battery, car adapter, upgradable

The 21264 machines start at $7119, with 128 MB RAM and 4.3 GB disk. That's with Windows NT, but you can always return that for a refund, right?