Linux On Big Iron

panker writes "eWeek is running an article about a company who converted their IBM mainframe into a Linux email server. "The technical support manager at Winnebago Industries Inc. recently oversaw the deployment of Version 7 of SuSE Linux AG's Linux operating system on an IBM zSeries mainframe to run his company's e-mail server supporting 700 users." "

Comments

So how well does it run?

[farsighed]

As someone who oversees an email system with close to a million addresses (in various states of utilization from heavy to "what account?"), I'm kinda interested in how this works out. Has anybody done high-load stresstesting of these yet?

-- F.S.

Re:So how well does it run?

[Anonymous Coward]

What - are you a sysadmin at hotmail?

Re:So how well does it run?

[WasterDave]

Not very well apparently:

"It's using between 7 percent and 10 percent of those to run Linux and the e-mail system."

For seven hundred email addresses, this strikes me as piss poor. Hell, a second hand P200 with good SCSI drives could run 700 accounts of qmail without breaking into a sweat.

Dave

Overkill???

[MadCow42]

You can support 700 users on a decent dual desktop system with Linux, what's this guy thinking?

Maybe he's just got an extra mainframe laying around...? You've got to think the support/maintenance on a mainframe would be horrendous compared to buying a new desktop server for this?

MadCow.

Re:Overkill???

[Karl Cocknozzle]

You can support 700 users on a decent dual desktop system with Linux, what's this guy thinking?

I would tend to agree, but it's also possible they're planning a large acquisition and need the horsepower to support a few thousand new employees instantly.

In these crazy days of "Merge merge merge" you never know...

Re:Overkill???

[Havokmon]

I would tend to agree, but it's also possible they're planning a large acquisition and need the horsepower to support a few thousand new employees instantly.

That's possible too, but head on over to Matt Simerson's FreeBSD Toaster. I'm SURE that could easily support a few thousand users in a clustered environment (NFS & Mysql)
Need more users? Add another box.

Re:Overkill???

[Garg]

It would be overkill if they were using the mainframe only for this. But they already had a mainframe. With these machines, you can create a logical partition and run Linux in it. That's what they're doing... just taking a few resouces away from their regular system.

Garg

Re:Overkill???

[zio pera]

Well, it has to be a very very big acquisition... A couple of year ago I set up a qmail server for handling the mail of a local ISP with about 30000 users (that's right, 30 thousands), and we used a dual cpu DELL poweredge, with only one cpu installed. The idea was to see if the box could handle the load, and eventually install the second CPU. Well, the load never got above 0.30, so we saved the money for the second CPU. And we are speaking of a PIII-500, no more.

Re:Overkill???

[ahde]

Set your mailbox limit to 4 megabytes. And put a link in your "message too big" folder that directs them to a tutorial on FTP.

Unless, of course, your service agreement states that you provide unlimited storage for every user.

Re:Overkill???

[ahde]

I haven't read it, but if this is the same Winnebego that switched like a year ago, I believe its 150K over 5 years.

Re:Overkill???

[Waffle Iron]

But they already had a mainframe.

I wonder what they did with the applications that used to run on that mainframe. I predict we'll find out next week's story on Slashdot: "Company Moves Business Applications from Monolithic Mainframe to Dozens of x86 PCs"

Re:Overkill???

[Garg]

They didn't do anything with the old apps. That's what I meant by "logical partition". You essentially create an address space that thinks it's its own machine, and can run a different OS.

The old stuff still runs without modification, albeit with slightly less resources.

Garg

Re:Overkill???

[tryavds]

Actually, the mainframe can be sliced into logical partitions, LPAR. My guess is that they already had a mainframe which they use with one of the mainframe OS's to run their accounting, inventory, or manufacturing. They allocated a small LPAR for this, put Linux and the mail server software on it and voila, no more exchange. The software they use (Insight Server from Bynari) allows running exchange type capabilities from outlook, so they replaced MS Exchange (which for 700 accounts is something like $50,000-$75,000 including OS, mail server, hardware for at least 2 intel boxex, CAL for both OS and Exchange), maybe more and not including Sysadmin, with this.

Re:Overkill???

[ahde]

So they spend thirty seconds learning which button to click. Your employees time is better spent learning proper resource management than watching movies on company time anyway.

And the discussion was about ISP users, not companies, which, by the way, already have size limits. If you're using POP3, it makes a whole lot of sense to cut out as many extra copies as you can. That's why MX was invented.

If one user is sending a 5 MB file to another, the likelihood is that the first user downloaded it from somewhere else (like a movie trailer), and can just send a link; or that it may be sent to multiple users, in which case it should be put on a shared drive, NFS, or some other networked partition, in which case, again, you can send a link.

Integrating FTP or some P2P file transfer mechanism into an email client might be a nice touch and would definitely cut down on the wasted bandwidth of multiple copying, but introduces more potential security risks than benefits.

Re:Overkill???

[Inoshiro]

Mail over NFS?

Don't do that.

Re:Overkill???

[Havokmon]

Don't do that

Not only do I think you're WAY off, but I don't think you read the toaster provided. Your link refers to Cyrus' IMAP server.

From the text:
Reliable: qmail's straight-paper-path philosophy guarantees that a message, once accepted into the system, will never be lost. qmail also optionally supports maildir, a new, super-reliable user mailbox format. Maildirs, unlike mbox files and mh folders, won't be corrupted if the system crashes during delivery. Even better, not only can a user safely read his mail over NFS, but any number of NFS clients can deliver mail to him at the same time.

From that page, the jist is: NFS isn't an issue for Maildir mailboxes.

Recycling

[SuperCal]

I love the idea of recycling old high-end machines. I think this could be a big thing in the future... Buying a cheap old mainframe could be cheaper and more reliable then using a PC workstation with server software. Besides it sounds like fun hehe.

Re:Recycling

[ostiguy]

It could be. But probably not.

We have a HPUX box in a 30ishU rack that is a big waste of space in our lab area. It would draw tons of power (you need a 30amp 110v twist lock socket for it, IIRC), it has raid - a raid of 2 gig disks, and its hpux isn't y2k compatible - it accepts 70-99 as valid years, IIRC, and onlyhas 128 megs of ram.

Honestly, for general unix stuff, mirrored 7200rpm drives and intel hardware would probably draw a fifth of the power, and be at least as quick (I would think it would be orders of magnitude quicker). I haven't bother pricing out HPUX upgrades for this nightmare. It just isn't worth the time or effort.

ostiguy

Re:Recycling

[jedidiah]

How is this different from the electric bill and cooling costs of a room full of glorified desktop XT's.

Harris County (Houston, TX)

[totallygeek]

I recently went to the Red Hat Certification class and there were two people from the core data services for Harris County. They were there to get some training on Linux because the county is deploying Linux on their IBM mainframes. And, yes, that means more than one!

By the way, if you are thinking of taking the RHCE course 300 (fast track for UNIX proficient people), I really enjoyed it.

Linux on a Mainframe

[saveth]

It's frequently said that Linux is not stable enough for mainframe systems. It's also frequently said that mainframes have been obsoleted by smaller, more powerful computers.

I am quite relieved to see that Winnebago has challenged the "norms" and put Linux to use on a mainframe. Linux is commonly used for mission-critical software, like the Linux server sitting next to me that handles our company's mail, but to see it doing something mission-critical on a mainframe is quite impressive.

Good work, Winnebago.

Linux under VM

[rlp]

IBM has a Virtual Machine OS, that allows you to run multiple OS's on a mainframe. You can run Linux (or even multiple instances of Linux) and still run your legacy apps under OS/390.

MIPS??

[Corby911]

"What's more, IBM officials said 11 percent of the mainframe mips (million instructions per second) shipped by the company in the fourth quarter of last year were configured to run Linux."

Perhaps it's just me, but that makes no sense whatsoever. How do you ship a measure of speed? "Shipping" millions of instructions per second seems to me to be the same as "shipping" miles per hour. It just doesn't make sense. My guess is that the author of the article got some terminology wrong.

Re:MIPS??

[Havokmon]

IBM officials said 11 percent of the mainframe mips

Since IBM machines can run paritioned (one machine, multiple os's at the same time), think of it as each machine IBM ships runs 11% Linux, the other 89% could be anything from OS/400 to AIX..

Correct me if I'm wrong..

Re:MIPS??

[owlmeat]

I don't have a problem with it. If you need an automotive analogy, compare MIPS to horsepower and it makes more sense.

Re:MIPS??

[CornfedPig]

more like shipping horsepower; it's a way of coming up with a single metric that normalizes the range of machines/configurations the company sells -- kind of like a capitalization-weighted stock index; the big boxes count for more, which they wouldn't if you just went by units shipped.

Re:MIPS??

[shakah]

Imagine that IBM sold three models, with the following MIPS capacity:
Model A: 10 MIPS
Model B (Linux): 100 MIPS
Model C: 1000 MIPS

And they sold the following amounts:
Model A: 5000 units (total of 5000*10=50000 MIPS)
Model B: 500 units (total of 500*100=50000 MIPS)
Model C: 50 units (total of 50*1000=50000 MIPS)

Don't you think a statement like "33% of our sales (by MIPS capacity) was configured for Linux" is a little more informative (and accurate) than "9% of our sales were Linux servers"?

Re:MIPS??

[WinPimp2K]

Big Iron is generally measured in MIPS, (been that way for over twenty years). So what the Big Blue Spokesperson is saying is:
"In the fourth quarter of last year, eleven per cent of the total computer power we shipped was tunning Linux."

Now that might mean that they shipped a total of 100 Mainframes (Really Big Boxes) of various models. They added up the MIPS of all of them and came up with some number of total MIPS -lets say 100,000. Of that 100 mainframes, thrity of them (relatively low end) totalling 11,000 MIPS were configured with Linux.

I'll agree, it seems kinda dumb, it would be like Ford reporting sales based on the total horsepower of all the engines in all the cars and trucks they sold, and then giving the percentage of them "configured" for diesel.
Maybe an UBMer would care to explain why it makes sense?

Re:MIPS??

[finkployd]

When you buy a mainframe from IBM you pay for the processing you want. They ship you a complete system (anywhere between 4 and 16 processors I believe?) and turn on (and charge you for) only the number you want. Some of the processors can be configured to only run Linux, and that is what they mean by "mips" being configured to run Linux.

The reason for configuring the processors to only run linux is that many mainframe software vendors charge based on the number of MIPS your machine runs at. If I am buying an MVS application I do not want my Linux processors in the mainframe to affect the price I am paying.

Finkployd

Re:MIPS??

[foobar104]

Perhaps it's just me, but that makes no sense whatsoever. How do you ship a measure of speed? "Shipping" millions of instructions per second seems to me to be the same as "shipping" miles per hour.

It's just you. ;-)

In the mainframe world, where virtualized hardware is the norm, systems are sold by the MIPS. In other realms of computing we talk about a 32-processor or a 128-processor system (or, if you buy Sun, a 106-processor system, for some reason). In mainframe land, you talk about a 12 MIPS system or a 45 MIPS system or whatever.

It makes a lot of sense, too, when you think about the fact that a four processor system, mainframe or otherwise, from five years ago is probably less powerful than a one-processor system now. It sort of normalizes it if you talk about system capacity in MIPS rather than in terms of "x processors of type y at z megahertz."

Don't compare the practice to shipping "miles per hour," but rather to horsepower. General Motors could, if they wanted to, say that they shipped umpteen million horsepower worth of engines last year, and it wouldn't be that hard to understand. It's just a different way of counting.

Re:MIPS??

[finkployd]

I'll agree, it seems kinda dumb, it would be like Ford reporting sales based on the total horsepower of all the engines in all the cars and trucks they sold, and then giving the percentage of them "configured" for diesel.

IBM charges you for your mainframe (and support contract) based on MIPS. The software you run in MVS (in most cases) is priced based on the number of MIPS you run in your mainframe shop. Ford doesn't charge for their cars based on horsepower :)

It makes sense because it lets you know that 11% of all mainframe computing power sold in the 4th quarter of last year is running Linux. IBM doesn't sell "boxes" per se, they sell MIPS.

Finkployd

Re:MIPS??

[ichimunki]

Because simply reporting unit sales doesn't accurately represent how much of the world's data is being processed by what OS, probably. As to the Ford example, I'm not sure they would do this by fuel type, but if they did it wouldn't be HP they reported, rather hauling capacity. Now if Ford had a stake in diesel fuel (like IBM seems to be taking in Linux), then you probably would see ads like this to make it look like Ford & diesel fuel were truly good options. Internally, this allows the business analysts to understand whether to devote resources. If I'm IBM I don't want to devote R&D time to Linux or any other OS simply based on machines running it, or licenses sold, I want to know how many customers are using it, and how much.

Re:MIPS??

[Corby911]

Ah, it was just a question of phrasing. The sentence I quoted was worded in such a fashion that I was confused. I fully understand the concept of MIPS and why it is often a better indicator of server capacity, I just read the text, and took the phrase "11 percent of the mainframe mips (million instructions per second) shipped" as "11 percent of the millions of instructions per second shipped", which makes a great deal less sense if you don't work with mainframes routinely(read: never). I probably would have put a footnote or something in the article explaining that, or tried to find a clearer wording (I've been trying for 5 mins or so, without luck).

(Thanks to all for the clarification)

Re:MIPS??

[brer_rabbit]

Maybe an UBMer would care to explain why it makes sense?

IBM, UBM, we all BM.

Re:MIPS??

[morcheeba]

One reason: Ford doesn't have to deal with the orders-of-magnitude in product variations that occur in computers, so counting vehicles means more than counting computers.

The limits of vehicles seem to be about 10:1 at most (2 passenger car : 50 passenger bus), 50:500 hp, maybe 1 ton car: 5 ton bus.

In computers you've got a much wider dynamic range... 1 user:1000 user, 32kHz (Sega VMU):2.2 GHz (P4), 1K:16GB memory, 56kbps modem : 2 Gbps fibre channel, etc.

I thought so too...

[WinPimp2K]

so I Read The Fine Article:

• he saved over 120K bucks vs an MSFT Exchange Server solution (for 700 users - what is Bill G smoking?)
• It is using less than ten per cent of the mainframe's capacity, they do plan to migrate other server jobs currently on discrete machines to the mainframe
• Yes they are already a mainframe shop. One mainframe is far more capable than a desktop server. Given a choice between supporting a single mainframe or a couple of hundred desktop servers, the mainframe costs less
• The article covers a lot more than just this one instance, but it has a nice number to start with - that big dollar savings

Re:I thought so too...

[TonyZahn]

what is Bill G smoking?

Rolls of hundred dollar bills, that's what.

Re:I thought so too...

[Trekologer]

It is using less than ten per cent of the mainframe's capacity, they do plan to migrate other server jobs currently on discrete machines to the mainframe

Exactly.

This sroty is turning into a whole bunch of "But I can do email suing this old 486 in my basement". But you are all missing the point. Since the mainframe's capacity isn't full, there are a lot of things that can be put on to it.

I guess no one saw the the IBM commercial where the company manager calls the cops because he thinks that all the servers were stolen. The tech walks out and says that they weren't and the room of servers was moved onto one machine. "It will save us a bundle."

Bingo.

Re:I thought so too...

[swb]

Is 120k for the licenses?

We have E2k for 500 users w/antivirus on a dual PIII 933 box w/2GB of RAM and the CPU meter seldom bleeps about 10% util.

They already had the frame

[drew_kime]

The need to ensure uptime was one reason Winnebago chose to run Linux on its mainframe. Last year, faced with an expensive upgrade to Exchange 2000, IT managers at the motor home and recreation vehicle manufacturer decided against the move and instead proposed running their e-mail system on Linux on the company's mainframe.

The company had success using Linux for domain name servers, Web serving and file sharing on its IBM S/390 mainframe running the Virtual Machine/Enterprise System Architecture operating system. After Winnebago officials decided they wanted their e-mail system on a reliable system, they chose to upgrade the company's mainframe, adding a second processor using IBM's virtualization technology, zVM, to run several Linux servers on a single mainframe.

They had the frame already, and just moved a new app onto it.

Overkill? Not at all.

[Samarkind]

There are several posts that are wondering if this is overkill, so I'll respond to all. It's not since they already had the hardware and only added a single CPU to their existing mainframe. They got the whole nine yards for $26K, but they don't have to add a new server, license Exchange, hire a Windows admin if they don't already have one and, as the exec said, they don't want to use Intel hardware.

We don't seem to be the target audience here...

[glwtta]

"In other words, Linux has come a long way from being Linus Torvalds' student computer project. "

Oh shit, it has? I better get on that there Linux thing then! Is it just me, or did that whole article have the "this isn't going to teach me anything new" feel to it?

Re:We don't seem to be the target audience here...

[zio pera]

"...in 1983 Microsoft made an equally risky decision to name its product after a term commonly used in the trade..." wich term ? DoS ? Sorry... could't resist ;)

Re:A mainframe for 700 users?

[swillden]

My company is planning to purchase a quad Xeon CPU 4GB RAM server as part of our Exchange 2000 migration for over 1000 users at a fraction of the price.

How is buying a new machine, plus a bunch of commercial software, plus sending techs to classes to learn how to manage the software, going to be cheaper than free software that your people already understand on a machine you already have? They did turn on an additional CPU, so there's some cost there, but more than likely they were getting close to maxing out their current system anyway and will benefit from the additional horsepower.

Plus, your solution is going to be far, far less reliable. There are IBM mainframes that have been running continuously for decades, with no downtime at all even through hardware and OS upgrades. Not that e-mail really needs to have six 9s uptime, but if you can get it for no additional cost, why not?

Happened 9 months ago...

[hangdog]

...according to this arcitle on IBM's Website.

Re:Happened 9 months ago...

[ahde]

Thank you. I was looking for that.
More links:

The Bynari site:
http://www.bynari.net/

From Consulting Times:

http://consultingtimes.com/connector.html

While we need case studies to show the actual savings that can be achieved through the deployment of InsightConnector, it's hard not to start counting additional money in the bank. For starters, the Connector fee schedule goes from $39 for a single user, down to $25 per seat for 100 users, with negotiable volume discounts for larger organizations. By contrast, Microsoft charges $92 per client licensed Exchange seat, so that a firm with 100 seats will experience a gross savings of $6,700, or 72 percent, right off the bat.

Linux using ESCON?

[Steve72]

How does Linux for mainframe connect to mainframe DASD? Do you have device files for 3390-3 and 9's? Do they mask it as SCSI volumes? What about FICON?

-Steve

Re:Linux using ESCON?

[spudnic]

So how much does an entry level z800 cost? I know the number would be wildly innaccurate, but say something to replace about 40 servers performing mail, web, dns, dhcp, etc.

Any ideas?

.

Re:A mainframe for 700 users?

[finkployd]

But if you already had the mainframe, it would be cheaper to just utilize it.

Finkployd

Re:A mainframe for 700 users?

[alen]

Forgot to mention. The Exchange databases will be stored on one of our EMC SAN's. In fact were looking at booting the whole thing from our Symettrix. And we just purchased a company with a Lotus Notes cluster on NT Enterprise. My guess is soon it's going to be a Win2000 Advanced Server cluster running Exchange 2000 Enterprise.

We have a Win2000 cluster running a Java app. All the problems come from the Java app.

Re:Overkill? Not at all.

[debaere]

of course, if I finished reading your reply (or bothered to read the article in the first place :) I'd have noticed the "Doesn't want to use Intel" bit, which makes my argument a tad pointless. oh well

It seems more appropriate now.

Re:A mainframe for 700 users?

[redhatbox]

Ah... but where would they get the nifty calendar stuff Exchange gives you? Oh, wait... we were talking about *email* servers ;). I never did get that stuff...

Sounds like a tremendous waste...

[jlockard]

What a complete waste. REALLY. I've got an old Dell server running an oldish version of RedHat and Cyrus, serving email for about 1200 users. The machine is far from taxed right now. I would say that the whole setup cost about $4000 when first purchased.

We're replacing that setup with a newish Dell 1U server running a newer version of RedHat and a newer version of Cyrus and making tweaks along the way. With being a school that has promised email accounts for life to alumni, we're planning for growth, but the server still cost around $4000 plus the cost of the RAID for email data store.

Re:Sounds like a tremendous waste...

[LinuxHam]

I've got an old Dell server running an oldish version of RedHat and Cyrus, serving email for about 1200 users. The machine is far from taxed right now

I think a lot of Intel-oriented people would be floored if they learned about the hardware reliability and flexibility that mainframes offer. Do you have dual power supplies in your Dell box? Can you hot swap them? Do you have hardware RAID? What about redundant hardware RAID attached to a dual-channel RAID storage box (also with dual hot-swap power supplies)? Can you hot plug your processors? RAM? RAID controllers?

A key feature of Linux is that it lets you select reliability and availability just by turning a dial. From handhelds to Intel to RISC to midrange and mainframe, you get to decide how mission critical your apps are. If you accidentally unplug your Dell box, your users are SOL until it reboots.

Re:Sounds like a tremendous waste...

[afidel]

actually our compaq's have all the features except hotswap motherboard components, and I believe some of the newer ones allow hot add of ram (though not hot replace?). The pci slots are all hot swap, the drives are all hot swap, the power supplies are hotswap etc. And they still cost a very small fraction of a mainframe.

Re:Sounds like a tremendous waste...

[LinuxHam]

The pci slots are all hot swap, the drives are all hot swap, the power supplies are hotswap etc. And they still cost a very small fraction of a mainframe

You're right.. I too have worked with Compaqs that have all those features. The original poster said he was running "an oldish Dell". Of course, the full-featured Compaq boxes did come close to the $18k mark, which approaches the $26k Winnebago spent on upgrading the mainframe to support their new Linux operations. Add another virtual server, and their choice was much more cost effective. Add ten more virtual servers and its no question which choice was better.

I'll run my setup script to bring another virtual server online in about 5 minutes, and you'll start over asking for another $15k for YAB.

Re:A mainframe for 700 users?

[Slashamatic]

You are doing what MS normally recommend I guess and you have a couple of systems (at least) just for E2K server, you did remember that second system to replicate on did you?

In the end you have a couple of boxes that runs E2K.

Yes, that was a full stop.What these guys are doing is running Linux under a single VM instance. It will cost them serious money because Linux for these boxes isn't cheap. However, they pay only for the first instance at their shop.

I have gone through the price options on W2K Enterprise Server, Advanced Server and Professional with Exchange Server and so on. There is *no* way that we are not talking serious cash here and that is for s/w alone. MS recommends that you dedicate particular systems for certain functionality like E2K - which is great but this costs.

The end result here is that you say that your company is planning to purchase. Come back when everything is working and tell us how much it really cost.

LPAR / Logical Partitions...

[Shivetya]

On their minis (AS/400) they are known as Logical Partitions, which allow the machines to run Linux, Notes servers, and various versions of OS400, and rumored to eventually include other OSes...

Many times easier to support one machine that multiples... and its easier to execute a backup machine for it as well.

Re:LPAR / Logical Partitions...

[finkployd]

They are called LPARs in z/390 lingo as well. However, you are limited to the number of LPARs you can run (12 or 14 if I remember correctly). Most shops who run Linux in production (as opposed to just playing with it) run it under zVM which has no limit as to the number of virtual machines you can run. Well, the processing power would eventually becoming a limiting factor, but there is no hard coded limit.

LPAR is literally a stripped down VM that runs in the microcode.

Finkployd

Re:Why I think IBM sucks

[delcielo]

IBM has consistently given me the best support I've found anywhere. They're patient, competent, and skilled.

The CE's I've dealt with have all been very professional; and have displayed none of the behavior you saw.

In one case, I even had a tech from supportline call me back weeks after he'd already solved a problem for me, to tell me about an alternate method he'd picked up from one of the more senior people. He'd used it on another customer's problem, and called me back to let me know it was an option that I might prefer if it came up again. Now that's service.

Overkill

[ruvreve]

As a hack this would have been nice, but email accounts for only 700 users doesn't need the capabilities of a mainframe. A high-end workstation or a low-end server would have fit this role nicely.

Options exist for about $25,000, that mirror each other so when one server crashes the other machine immediately takes over.

Now for the MS Exchange licenses I don't know the exact cost but $100,000 seems extreme. If you have the support staff to run the linux email server on a mainframe then its probably a viable option. Otherwise, your going to pay more in support in the next few years.

Picking up steam..

[LinuxHam]

if you are thinking of taking the [fast track] RHCE course I really enjoyed it.

Same here, last June. And when I took my Linux on S/390 training in December, I was in class with people from a major online bill payment company, a major auto insurer, Canada's DOD, and many others. Most had already deployed it and wanted to see what they hadn't figured out for themselves yet.

In August, 2000, I sat next to an IBM'er by coincidence on a flight. He saw I was reading the "Linux for S/390" RedBook. He said I'd become a "demigod" if I get into that. I've already gone thru one consolidation project, starting a second one on Monday in NYC, and have a third one queued up, waiting for me to finish up in NY. It seems we recently gave a customer a server upgrade plan, and they replied, "what, no Linux?" So we're redoing it as a consolidation plan for Linux on S/390.

I'd say Linux on S/390 is picking up steam big time. When I spoke to a friend about this setup, he replied, "Wow, you finally sound like one of those mainframe IBM'ers we used to make fun of!" Of course, he still has no reply to the argument that I can reduce just about any single data center to a couple of 48U racks, and give all the servers five nines.

Re:Picking up steam..

[Telastyn]

My father now works for IBM (bought and outsourced to them) so he gets all the newsletters and the such. Last year IBM had something akin to 20x mainframe sales after the linux initiative. It pretty much saved the department.

The only problem I've seen is most current admins are used to/learned linux on little dinky spare desktop machines. 'Mainframe' carries a big scary connotation. The name itself intimidates, like a *nix prompt scares most MCSE's.

Plus most bosses 'know' that mainframes are *so* 1970's...

Re:Picking up steam..

[totallygeek]

It is great to see Linux (a free product) drive major sales in the computer industry for hardware and software solutions. I don't think that anyone thought it possible this soon if ever. I remember everyone I trained with, but there were five people from Compaq (well, there are in Houston), and they were all five vintage DEC people. There was a guy from Texas A&M.

I would have liked it when I was at A&M if we ran Linux.

Re:Picking up steam..

[Banjonardo]

Heh, your post just hit me with the full irony:

In the early 80s, IBM used MS-DOS because it thought hardware was where the money is. Now Microsoft is, well....Microsoft.

Now in the early 00s, IBM is using a free operating system and raking the big bucks in... you guessed it! Hardware!

Aahhhh, computer irony.

Re:You missed something

[Ooblek]

Nah, they are secretly planning to overtake all the extra bandwidth on the internet by processing high-volumes of spam. I think they code-named the mainframe MCP or something......it might talk too, but I'm not sure.

Winnebago is always on the cutting edge..

[asv108]

Ever see SpaceBalls?

Read Article?

[ackthpt]

700, yeah, ok, sure, there's about 200 posts all approximating: "What? I can support 700 mail users on a cruddy old 486" (which we once did, btw, it worked peachy)

But if you read the article you see that they don't want to support more servers, they want to support less, i.e. not buying anymore Intel servers, which are like so many cats. The only downside I see is a single point of failure, the zSeries goes 'poot!' and the staff takes the afternoon off.

But adding that function to an existing piece of hardware does keep support costs down, and as they've noted, pay once to get their mail running on there, as opposed to paying Microsoft for Exchange, year in, year out, well, it looks smarter, doesn't it?

Last, but not least, if they decide to move it off the mainframe later, hey, they should be able to migrate it with little pain, since the OS runs on just about anything.

'Bago 2005: "Tell eWeek we just moved the entire mail server to a hacked TiVo, will you, don't forget to mention it does voice and video email, too."

Re:Why I think IBM sucks

[rosewood]

I will say that the phone people and the first CE I delt with was a okay

they two frik and fraks kill me

Email != groupware

[einhverfr]

Exchange and Bynari are groupware packages which incorporate email functionalities within their capabilities. Same with PHPGroupWare.

If you just wanted email, you would probably stick with SendMail, Qmail, or the like. However, the shared calendars, todo lists, etc. are important in the groupware environment!

I feel compelled to quote Scott McNealy

[AnalogBoy]

Just where you can get sun microsystems's view on this (not neccesarily mine.. but perhaps.. but maybe not..)

To quote This Article in computerworld magazine:

Q: Sun has done quite a bit in the way of Linux support, but you really haven't gone the IBM route of marketing Linux-based systems. Why is that?

A: We're the No. 1 Linux appliance server supplier in the world with the Cobalt line [from the acquisition of Cobalt Networks Inc. last year] (see story). We have Linux extensions to Solaris. We just don't think a Linux partition on a mainframe makes a lot of sense. It's kind of like having a trailer park in the back of your estate.

Re:Overkill? Not at all.

[ahde]

There are a couple of things you're missing.

1) They're running IMAP, not just POP3 (though your system could probably handle 700 IMAP users easily)

2) They're using Insight from Bynari so that all their outlook clients can have their popups.

They may have antivirus software too, and that will really eat CPU and I/O.

Re:Overkill? Not at all.

[Steve Hamlin]

It's not since they already had the hardware and only added a single CPU to their existing mainframe. They got the whole nine yards for $26K, but they don't have to add a new server, license Exchange...

And that's what I think the earlier posters were talking about, and what still hasn't been answered.

It makes sense to spend $26,000 on a zSeries/Linux solution over "spending $150,000 on new hardware and software for a Microsoft Corp. Exchange upgrade."

But why spend to the $26,000 at all if you can support 700 users on a $5,000 semi-high-end traditional Linux x86 email server?

Is it worth the $26k to not have to worry about an extra physical box? The administration is the same, it is just running Linux on a zSeries for $26k vs. running Linux on a x86 SMP for $5k.

What am I missing? Does the $26,000 include a bunch of consulting services that they needed? $26k to have the IBM name and support?

Not a waste...

[DarkMan]

Take an AK-47

Take a magazine of 7.62 mm S ammo, and load.

Walk into computer room.

Aim at [Your Dell, The IBM]. Empty magazine at said server.

Which one still works?

More usefully, which one can you fix up back to full condition, without losing a single email?

That's reliability.

How a Mainframe works

[pcs305]

I note that there is quite a few peoples that do not know how things are done on a mainframe. I will TRY and clarify. If you have an average box, say 400 mips, there will be a couple of things happening on this box. Starting of with the OS, OS/390 (or z/OS, VSE, VM). Then a network server VTAM and TCP/IP, a Security server, the some DB's DB2, IMS/DB(or some vendor DB). Then transaction servers CICS or/and IMS serving 4000+ concurrent online users logged on to CICS or IMS. Then you may have oh lets say 500+ programmers, system programmers, DBA, Administrators etc, logged on to the OS "shell" TSO doing programming, compiling, admin, editing and in general doing what these kinda people do, maintaining the monster. The there may be a MQSeries or two running handling client connections and messaging applications and client (pc's, server's *nix's) connections to DB2.. (i'm touching the surface here!) So in general there is a little more happening on a mainframe than on you average wintel box. So in order to separate the production and development, and test environments you can go and partition this one little box up into three logical partitions, called LPAR's in dinosaur speak. Each LPAR can be IPL'd (BOOT in dino speak) without affecting any one of the other LPAR's. So now you have one IBM 2064 400MIP box, 3 LPAR'S and still have room to breathe. Add another LPAR, install VM, Load LINUX/390 and reboot the LINUX LPAR. But still only one Linux server on the box? But we have 250 NT boxes to replace? Not to Worry!! under VM on your brand new LPAR on your mainframe box you can begin to start Linux images at will. So in the end we have 1 Box, 3 OS/390 Mainframe partitions, 1 VM partition with 250 Linux servers running. But now the 400 mips is kinda running out of steam. Call up IBM and they will come and add another cpu. And if you were planning ahead the IBM 2064 should have an idle couple of CPU's under the hood not being used. So you call IBM, they give you the code and you go to the master console and issue a "very cpu online" command and off you go. No IPL required, no downtime. Do that on your duel XEON Intel with Windows... any kind of Windows. I'm not going to go into Parallel Sysplex'ing, syscon's and CICSPlexing etc.

Oh puh-lease I assume this is a joke???

[coupland]

Two points of contention here folks:

1. Linux on big iron
You're telling me that running a 700-user e-mail system on Linux is an example of Linux on big iron? Sure the server is big but I could do the same job on a medium-range Intel workstation. Are we supposed to be impressed by this feat of server load balancing? Whooo.... 700 POP accounts on a single server, it's magic...

Second point:

>the move allowed him to avoid spending $150,000 on new hardware and software for a Microsoft Corp. Exchange upgrade

$150,000 for a 700-user Exchange infrastructure?!?!? Where does he work cause I thought my company liked to throw money into the wood chipper! You could easily support 700 users with full redundancy for $60k. Ever seen FUD working in the opposite direction? You have now.

Re:I feel compelled to quote Scott McNealy

[jedidiah]

Actually, Unix is supposed to be architected specifically to allow the trailer trash and the bluebloods to next to each other in perfect harmony. Is this Sun's way of admitting that they don't design their Estates well enough to accomodate a wide variety of occupants?

Although, the key thing that is not being said here is: Next to zOS, Solaris is just a "double-wide".