Slashdot Mirror


IBM Launches Linux-Only Mainframes

An anonymous reader writes: IBM is introducing two mainframe servers that only run on Linux. It's part of a new initiative from the Linux Foundation called the Open Mainframe Project. "The idea is that those companies participating in this project can work together, and begin building a set of open source tools and technologies for Linux mainframes, while helping one another overcome common development issues in the same manner as all open source projects." IBM's hardware release is accompanied by 250,000 lines of code that they're open sourcing as well. "Ultimately the mainframe mainstays are hoping to attract a new generation of developers to their platform. To help coax new users, IBM will be offering free access to the LinuxOne cloud, a mainframe simulation tool it developed for creating, testing and piloting Linux mainframe applications." Canonical is working with IBM to bring Ubuntu to mainframes.

157 comments

  1. In Capitalist America... by wonkey_monkey · · Score: 2, Funny

    IBM is introducing two mainframe servers that only run on Linux.

    In Capitalist America, Linux runs on mainframe servers.

    --
    systemd is Roko's Basilisk.
    1. Re:In Capitalist America... by Anonymous Coward · · Score: 0

      Hurr hurr hurr.

    2. Re:In Capitalist America... by Lunix+Nutcase · · Score: 1

      Was that supposed to be funny?

    3. Re:In Capitalist America... by GNious · · Score: 1

      Did it have to be?

    4. Re:In Capitalist America... by MobileTatsu-NJG · · Score: 1

      Yes.

      --

      "I like to lick butts!" by MobileTatsu-NJG (#32700246) (Score:5, Informative)

    5. Re:In Capitalist America... by WindBourne · · Score: 1

      I think that you mean, capitalist India. Right?

      --
      I prefer the "u" in honour as it seems to be missing these days.
    6. Re:In Capitalist America... by wonkey_monkey · · Score: 1

      Yes. Was that ^ supposed to make you look superior?

      If you didn't get it, or did and just didn't think it was funny anyway, that's okay. But what makes you care so much that you felt you had to tell everyone that you didn't find it funny?

      At least one other person has found it funny enough to mod it so, so I'm calling that a win.

      If I've misinterpreted your post, and you're upset because you recently lost someone dear to you in a Linux mainframe related accident, then you have my sympathies.

      --
      systemd is Roko's Basilisk.
    7. Re:In Capitalist America... by K.+S.+Kyosuke · · Score: 1

      Was that a sound of a DASD unit actuating its big heads using powerful stepper motors?

      --
      Ezekiel 23:20
    8. Re:In Capitalist America... by Anonymous Coward · · Score: 0

      What about 'Funny Adjacent'? Worth a smirk and quiet chuckle but not quite RATFLOL ...

    9. Re:In Capitalist America... by Anonymous Coward · · Score: 0

      Well ... they are not quite 'Lost', so to speak. The process is there and consuming space but not really doing anything worth while except for blocking the job queue.

    10. Re:In Capitalist America... by MobileTatsu-NJG · · Score: 1

      Yes, if it had reached that level it would have sufficed.

      --

      "I like to lick butts!" by MobileTatsu-NJG (#32700246) (Score:5, Informative)

    11. Re: In Capitalist America... by Anonymous Coward · · Score: 0

      'A'? At? Above? Along?

  2. What about Fibre users? by Anonymous Coward · · Score: 5, Funny

    To help coax new users, IBM will be offering free access to the LinuxOne cloud,

    Is this access just for coax users or is it available via fibre or twisted pair?

    1. Re: What about Fibre users? by Anonymous Coward · · Score: 0

      Thanks for your inquiry. Our generous offer is available to first time coax users only.

    2. Re:What about Fibre users? by silas_moeckel · · Score: 2

      Twinax, IBM's old 5250 cable is all the rage again, running new sas/sata displayport and 10-100ge.

      --
      No sir I dont like it.
  3. Been around for awhile... by perotbot · · Score: 1

    Suse has supported linux on Z-Series for quite some time https://www.suse.com/products/...

    --
    ~corporate tool, but employed~
  4. The only catch by Anonymous Coward · · Score: 1

    You'll need an EBCDIC keypunch to create the input card deck.

  5. THIS is the kind of story... by LVSlushdat · · Score: 1

    THIS is the kind of story /. was designed for..... VERY cool!!

    --
    THANK YOU, Edward Snowden!! Americans owe you a debt of gratitude (whether they know it or not..)
    1. Re: THIS is the kind of story... by Anonymous Coward · · Score: 0

      No. Slashdot is a nerd site, not an IT site. It is an interesting topic for nerds, but not definitively so.

  6. Ubuntu?! by Anonymous Coward · · Score: 2, Insightful

    I'm sorry, but Ubuntu on mainframes? Ubuntu is the linux distribution FURTHEST from being appropriate for a mainframe - it's heavily targeted towards desktop users, particularly those with a lower level of expertise (or a lower desire to put work into their OS) than the average linux user. What's more, it's adware/spyware now, which is definitely something I'd hate to have on a mainframe - the last thing you want is your OS transmitting and receiving data at random!

    1. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      Yeah, perhaps they mean ubuntu server or something...

    2. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      You know you don't have to install all of the packages, right?

    3. Re:Ubuntu?! by Dareth · · Score: 2

      Ubuntu is a popular Linux distribution among developers. While I personally prefer Debian, there are plenty of people using Ubuntu to do some very interesting work.

      --

      I only look human.
      My mother is a halfling and my dad is an ogre, so that makes me an Ogreling
    4. Re:Ubuntu?! by silas_moeckel · · Score: 1

      None of which is realy to go onto big iron.

      --
      No sir I dont like it.
    5. Re:Ubuntu?! by GCsoftware · · Score: 1

      I've run a few Debian/s390x instances in Hercules and it's pretty cool. Once it's up and running, you wouldn't know you're running on a "mainframe"

    6. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      If they tasked me with deploying Ubuntu to a mainframe by the end of 2018 I'd quit right now.

      Ubuntu... used by windows users who think it's Linux.

      that said.... It's nice to see the forward thinking here, sure we've got solutions in place but this can lead to some very interesting projects in the next few years.

    7. Re:Ubuntu?! by Blaskowicz · · Score: 1

      That's like saying Debian and Red Hat are for illiterate cell phone users because they have GNOME 3.

    8. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      Which is just a different set of packages. In the end Ubuntu is just a package archive.

    9. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      Noone cares what flavor of Linux it runs. It's your program that's important, plus the DB it uses. If your program compiles with no warnings with gcc then you're good on any linux.

    10. Re:Ubuntu?! by RabidReindeer · · Score: 2

      If they tasked me with deploying Ubuntu to a mainframe by the end of 2018 I'd quit right now.

      Ubuntu... used by windows users who think it's Linux.

      that said.... It's nice to see the forward thinking here, sure we've got solutions in place but this can lead to some very interesting projects in the next few years.

      Oh come on. I'm mainly a Red Hat person, but, excepting Suse, there's no other Linux distro that's more server-friendly than Ubuntu, whose work in that area goes back a decade.

      Right now, the cutting-edge work on clouds and containers is mainly being done on Red Hat and on Ubuntu. And sometimes the more advanced stuff is coming from Ubuntu. If Ubuntu hadn't also corrupted itself with systemd, I'd be seriously considering converting a couple of RHEL boxes.

    11. Re:Ubuntu?! by shadowknot · · Score: 1

      I spoke with some guys running the Ubuntu booth at last year's IBM Enterprise conference in Vegas. They were there to tout their System p distro and when I quizzed them on the potential of a z port I got the deer in the headlights, what are you talking about look. Now that could've just been the guys I was talking to and there may well be some z enthusiasts back at Shuttleworth Towers but from my experience they really didn't seem interested. If you're really serious about running Linux on z you most likely run SUSE (SLES), Red Hat is actually rather behind on the platform.

    12. Re:Ubuntu?! by Junta · · Score: 1, Interesting

      you wouldn't know you're running on a "mainframe"

      Your accountant would know if it were really running on a mainframe. Your users and developers wouldn't notice any benefit, but your accountant can painfully feel the weight of the mainframe.

      This is IBM grasping at relevance of their mainframe platform to a wider audience. The problem is that it's not an appealing architecture for those workloads. If anything this may be making some hardcore mainframe shops wonder more strongly if they should be moving off, since even IBM seems to be legitimizing the 'not-mainframe' way of doing it for mainframe users. Along the lines of how OS/2's windows compatibility made it so that workloads shouldn't bother targeting OS/2 since IBM would support it via windows compatibility. IBM hoped windows compatibility would make users prefer OS/2 so they could run both, but it backfired on the developer end.

      --
      XML is like violence. If it doesn't solve the problem, use more.
    13. Re:Ubuntu?! by Anonymous Coward · · Score: 0

      I'm sorry, but Ubuntu on mainframes? Ubuntu is the linux distribution FURTHEST from being appropriate for a mainframe - it's heavily targeted towards desktop users...

      You don't seem to be aware of Ubuntu Server http://www.ubuntu.com/download/server/

    14. Re:Ubuntu?! by jellomizer · · Score: 1

      Isn't that just Debian?

      I don't see any real advantages of using Ubuntu server over Debian? Is there any?

      --
      If something is so important that you feel the need to post it on the internet... It probably isn't that important.
    15. Re: Ubuntu?! by Anonymous Coward · · Score: 0

      I thought RH started using systemd as well.

    16. Re:Ubuntu?! by rubycodez · · Score: 1

      Eh, there is not much difference between Debian and Ubuntu Server, and Debian was the best server Linux distro until the systemd twats ruined parts of it the same as the other major server distros

    17. Re:Ubuntu?! by rubycodez · · Score: 1

      You are spewing without knowledge, mainframe the most cost effective solution for many enterprise workloads

    18. Re:Ubuntu?! by rubycodez · · Score: 1

      You are confused, Ubuntu Server is inferior how from Redhat or SLES? Since I admin hundreds of servers running all of those (plus Debian and CentOS), I can definitely tell you in what areas Redhat/CentOS and SuSE are inferior

    19. Re: Ubuntu?! by rubycodez · · Score: 1

      Yes it does, RedHat are major contributors to that bloated rubbish. Sadly all the distros that are "enterprise certified" for various common wares are going to systemd. A good argument for pressuring vendors to certify on BSD I say...

    20. Re:Ubuntu?! by Junta · · Score: 1

      I would suggest that the areas where it is the most cost effective solution to the problem at hand (or believed to be the most cost effective) have low overlap with linux applications. A z without z/OS is a rather silly overprised, under-capability system.

      --
      XML is like violence. If it doesn't solve the problem, use more.
  7. No it hasn't by drinkypoo · · Score: 2

    These are mainframes that only run Linux. There is no other supported operating system.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    1. Re:No it hasn't by erikscott · · Score: 3, Interesting

      Yup - first thought that ran through my mind: "Oh, they're selling Z Series with crippled Firmware."

      I'm kind of stumped. Linux on a Mainframe is a neat party trick, but it doesn't really make a lot of sense. Modern Z Series hardware is heavily derived from Power. Why not just run Power Linux? Mainframe I/O design is intentionally about as un-PDP-like as possible, so it's a bad match for Unix, Linux, or even Windows for that matter (NT ran on MIPS, so it theoretically could be ported to S/390). Mainframes get their performance by pushing computation into the channel controllers, and while you could do something like that in Linux, are any of your applications ready to treat your database like a device driver? Because that's what you'll have to do. And, incidentally, it's why every attempt from AIX/370 to Linux on Z Series has required virtualization and a ton of independent kernels to get anything resembling decent performance. And that's where Dell will come in and put thousands of cores in a 42U rack for you... No, IBM's own P Series is a better idea, and their former x86 division (now Lenovo) looks even better.

      Erik

    2. Re:No it hasn't by GCsoftware · · Score: 2

      Yeah, I'm sort of failing to see the point here as well - running your classic z/OS backend stuff and then having a few zIFLs talking to the backend over HiperSockets (IIRC) made sense, but just a big zSeries box with no way to run legacy apps?

      I'm not sure but I guess the market will decide..

    3. Re:No it hasn't by bws111 · · Score: 5, Informative

      Modern Z hardware has nothing to do with POWER. Mainframes do not push computation into channel controllers, whatever gave you that bizarre idea? Treat your database like a device driver? What is that supposed to mean? Linux runs native on zSeries, so virtualization is not necessary (and has not been for more than a decade).

      You seem to know absolutely nothing about mainframes, why are you posting?

    4. Re:No it hasn't by bill_mcgonigle · · Score: 3, Informative

      And that's where Dell will come in and put thousands of cores in a 42U rack for you...

      We're getting to the point where all that matters is how much performance can you get from an assemblage of nodes, and how much does it cost to buy and support it?

      If IBM can provide a lower TCO than Dell with different technology and the "containers" are compatible, many customers will be interested.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    5. Re:No it hasn't by bws111 · · Score: 1

      Linux-only mainframes (all IFLs, no CPs) have been available for years. This is not a new type of machine, it is more of a 'solution'. The major new thing is that you can use KVM to manage virtual images instead of z/VM.

    6. Re:No it hasn't by Anonymous Coward · · Score: 0

      Mainframe I/O design is intentionally about as un-PDP-like as possible

      Isn't it more accurate to say PDP I/O design is intentionally about as un-mainframe-like as possible? Since mainframe I/O design is based on principles established well before the PDP-11 (or even PDP-1) existed. Isn't there still software buried in the I/O stack that is kind of emulating a punch-card reader?

    7. Re:No it hasn't by Anonymous Coward · · Score: 0

      I'm also stumped. A mainframe's job is I/O, MIPS, and making sure every component is as reliable as possible, from the DASD on down.

      Yes, Linux works well on a VM, but as the hypervisor OS, it doesn't make sense. Things like CPUs running in lockstep and Parallel Sysplex are not items easily grafted onto an OS. It would be about as big a step as trying to make Linux work on a Harvard architecture where code and data are in separate address spaces in RAM.

      Yes, you can pound a screwdriver into a hammer, but of all the things that Linux isn't designed for, is a mainframe OS, because either the mainframe hardware will have to be changed to support Linux's "view" of the world, or Linux has to get a ton of drivers.

      I understand IBM's mainframe division is trying to stay relevant, but they need to focus on getting people to use a mainframe, not trying to make a mainframe act as a PC. The main thing is to work on the hypervisor, and focus on Linux as the guest OS, and let the hypervisor do all the background work, so the DASD which is presented as address space to a mainframe OS is presented as a virtual SCSI drive to Linux. Things like lockstep CPUs and such, leave that to the hypervisor, and let Linux view it as one CPU in /proc unless there are critical exceptions that need to be passed to the client OS.

      However, what really needs to be done is sell what a mainframe does best, and that is reliability. A company can buy a mainframe for bottom-up reliability, or throw lots of man-hours to code a backend like Facebook's so the hardware can be pretty much disposable. What IBM needs to do is have a case for having the hardware be expensive and reduce the number of man-hours needed to be put in to code a solution.

      This is where mainframes excel. You need at most a handful of mainframe operators once the machine is in place.

      Of course, POWER is a gem IBM needs to leverage. This is an architecture that Linux is meant for and would work well on. Even though IBM does not have the economies of scale Intel and AMD enjoy, if they focused on bang for buck and PowerVM, adding some features [1] so it keeps up with the MS side of the fence... and add emphasis on security. AIX, hands down, is quite a secure OS. Not 100%, but

      [1]: Something like MS's Storage Spaces Direct, with Infiniband connections between the POWER boxes that presents the backend hard disks as one volume, for example. That way, no SAN is needed, if done right, and disk space can be expanded by just adding more nodes.

    8. Re:No it hasn't by afidel · · Score: 1

      Thanks, I was thinking that there had been an L part number z series box for at least a decade and was wondering what the news was. In fact slashdot covered their first linux only part 13 years ago

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    9. Re:No it hasn't by Tough+Love · · Score: 1

      Linux on a Mainframe is a neat party trick, but it doesn't really make a lot of sense.

      That "party trick" is generally credited for reviving IBM's mainframe business.

      Modern Z Series hardware is heavily derived from Power. Why not just run Power Linux? Mainframe I/O design is intentionally about as un-PDP-like as possible, so it's a bad match for Unix, Linux, or even Windows for that matter (NT ran on MIPS, so it theoretically could be ported to S/390).

      Not sure what you're concerned about. Linux sees the mainframe hardware through drivers like any other architecture. Who cares if the DASD driver using a channel is organized differently from a memory mapped driver for some other arch? It just looks like a block driver to the rest of the kernel.

      Mainframes get their performance by pushing computation into the channel controllers, and while you could do something like that in Linux, are any of your applications ready to treat your database like a device driver?

      Even more puzzled by this comment. S390 applications never need to bother with channel controllers or know they exist.

      Because that's what you'll have to do.

      I doubt it.

      And, incidentally, it's why every attempt from AIX/370 to Linux on Z Series has required virtualization and a ton of independent kernels to get anything resembling decent performance.

      Citation needed.

      And that's where Dell will come in and put thousands of cores in a 42U rack for you... No, IBM's own P Series is a better idea, and their former x86 division (now Lenovo) looks even better.

      IBM covers the whole spectrum of form factors. For some customers, fewer individual boxes looks like a great idea.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    10. Re:No it hasn't by Tough+Love · · Score: 1

      Yeah, I'm sort of failing to see the point here as well - running your classic z/OS backend stuff and then having a few zIFLs talking to the backend over HiperSockets (IIRC) made sense, but just a big zSeries box with no way to run legacy apps?

      It seems apparent that there now exists a significant market segment that only cares about Linux serving and not legacy mainframe apps.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    11. Re:No it hasn't by Tough+Love · · Score: 3, Insightful

      ...of all the things that Linux isn't designed for, is a mainframe OS, because either the mainframe hardware will have to be changed to support Linux's "view" of the world, or Linux has to get a ton of drivers.

      Now I'm stumped by your comment. Linux has no problem with "a ton of drivers", but the fact is, only a few drivers are actually needed.

      I understand IBM's mainframe division is trying to stay relevant, but they need to focus on getting people to use a mainframe, not trying to make a mainframe act as a PC.

      What makes you think Linux makes a mainframe act as a PC? Linux runs on many disparate architectures, some of which look very little like a PC. Sure, Linux forces every architecture to present a page table abstraction derived historically from intel's model, but is that is mainly because that simple model makes sense, and not particularly inefficient for architectures with a different approach to emulate it. Other than that, life in mainframe land is much like any other architecture, especially now that with virualization rampant, everything looks a lot more like a mainframe inside today.

      Things like lockstep CPUs and such, leave that to the hypervisor, and let Linux view it as one CPU in /proc unless there are critical exceptions that need to be passed to the client OS.

      Why would you think it works any differently than that?

      However, what really needs to be done is sell what a mainframe does best, and that is reliability.

      How do you imagine IBM sells mainframes? There was a time when customers had a lot of idle capacity siting around that could be recycled as Linux servers, but now IBM's only compelling argument is reliability. But that argument is a cruncher for some customers.

      What IBM needs to do is have a case for having the hardware be expensive and reduce the number of man-hours needed to be put in to code a solution.

      The mainframe proposition is not about maintenance cost, it is about the business cost of even temporary interruption or failure.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    12. Re:No it hasn't by Anonymous Coward · · Score: 0

      Wow, no z/VM! That's a major selling point. The awfulness of z/VM is amazing. I'd buy one of these things just to get rid of z/VM forever.

    13. Re:No it hasn't by rubycodez · · Score: 1

      you have some incorrect information also

      First layer virtualization is provided by PR/SM to deploy LPARS prior to this announcement. Linux was not running "natively"

      Of course the Z architecture is modified PowerPC processor based

    14. Re:No it hasn't by rubycodez · · Score: 1

      Those "linux only" mainframes still had LPARS under PR/SM first layer hypervisor, this new machine will allow KVM along side of PR/SM

    15. Re:No it hasn't by bws111 · · Score: 1

      z processors are not and have never been based on PowerPC. I don't know where this myth comes from.

      When someone says that an OS running on a platform 'requires virtualization', they mean that there is some required facility missing from the platform, and that facility is provided by the virtualization layer. There is no such requirement to run Linux on Z. PR/SM is not providiing anything that is not available in the hardware.

    16. Re:No it hasn't by Guy+Harris · · Score: 1

      Of course the Z architecture is modified PowerPC processor based

      So you're claiming here that the uops into which the z13 microprocessor cracks the non-single-cycle z/Architecture instructions are Power ISA instructions?

    17. Re:No it hasn't by bws111 · · Score: 1

      It doesn't allow KVM along side PR/SM. Before this announcement you could have Hardware->PR/SM->Linux or Hardware->PR/SM->z/VM->Guests. Now you can also have Hardware->PR/SM->Linux->KVM->Guests.

      This announcement is not a new machine. The machine is z13. The announcement is the ability to virtualize with KVM instead of z/VM.

    18. Re:No it hasn't by bws111 · · Score: 1

      What exactly is so awful about z/VM?

    19. Re:No it hasn't by erikscott · · Score: 2, Informative

      Z series and power definitely do not share an instruction set, and they have really substantial differences, but that isn't keeping the engineering teams all that separated, if indeed they are at all.

      Quoting Timothy Prickett Morgan from http://www.itjungle.com/tfh/tf... , "And as has been the case in the past, the Power and z processors are designed by a single processing team and are borrowing technologies from each other. This does not, however, mean that IBM is creating a converged processor that can support either Power or z instruction sets." My hazy memory makes me think they're sharing FPU blocks, possibly one of the bus interfaces, and it seems like one of the cache blocks (L3?). Z has plenty of custom hardware - I think it's fair to say it's predominantly custom - the branch predictor would have to be pretty different, and of course power doesn't have a BCD arithmetic unit.

      Point being, if you're going down the Z Series road to run a Unix-like OS, why not just (conceptually) stop early, end up with something like Power, and call it good? Anyway, I'll argue that they're spiritually and economically related, and there's more than a passing family resemblance. Kind of like power and modern ("advanced server") iSeries, though that's getting more into Deliverance territory.

      Meanwhile, channel controllers aren't as dumb as they look. A little wikipedia action here (I know, citing wikipedia, but it's monday and I'm still tired): https://en.wikipedia.org/wiki/... . Turns out the little dickens can do a decent amount of work on its own. I think the wikipedia entry is showing its age... seems like IBM's done a lot more work since this.

      I remember when SASI came out. I looked at the spec and thought "Hey, this is a lot like a channel controller." Then I read some more and decided "No, a channel controller is much smarter. But this isn't bad." SASI became SCSI and everything else flowed downhill from that. At a very real level, Linux is forcing a million dollar fibre channel array to look more or less like an ST506 connected a board from 1984. Wild.

    20. Re:No it hasn't by bws111 · · Score: 1

      They are not 'going down the road'. Linux on Z, and the IFL (Integrated Facility for Linux) has been around for 15 YEARS. This is not something new. I don't know why you think there is some sort of incompatibility between mainframes and Linux.

    21. Re:No it hasn't by Guy+Harris · · Score: 2

      Z has plenty of custom hardware - I think it's fair to say it's predominantly custom - the branch predictor would have to be pretty different, and of course power doesn't have a BCD arithmetic unit.

      Actually, it does have IEEE decimal floating-point, as does z/Architecture. z/Architecture has decimal fixed point, but, these days, it might just trap to millicode doing tricks such as excess-6 for carry propagation. (And the PowerPC processors in at least some AS/400 machines added some instructions to assist BCD arithmetic.)

      Anyway, I'll argue that they're spiritually and economically related, and there's more than a passing family resemblance. Kind of like power and modern ("advanced server") iSeries,

      There is no iSeries any more, there's just the IBM Power Systems, which are the successors to both RS/6000^WpSeries^WSystem p and to AS/400^WiSeries^WSystem i; they can run both AIX and IBM i.

      Meanwhile, channel controllers aren't as dumb as they look. A little wikipedia action here (I know, citing wikipedia, but it's monday and I'm still tired): https://en.wikipedia.org/wiki/... . Turns out the little dickens can do a decent amount of work on its own. I think the wikipedia entry is showing its age... seems like IBM's done a lot more work since this.

      Yes, but they're still I/O boxes, not general-purpose computers (well, they might be implemented with z/Architecture or Power ISA processors, but what's exposed to the OS or application programmer is just the ability to run limited channel programs). The z/Architecture Principles of Operation says in "Execution of I/O Operations", in chapter 15 "Basic I/O Functions":

      For subchannels operating in command mode, the channel subsystem can execute seven types of commands: write, read, read backward, control, sense, sense ID, and transfer in channel. Each command, except transfer in channel, initiates a corresponding I/O operation.

      and

      For subchannels operating in transport mode, the channel subsystem can transport six types of com- mands for execution: write, read, control, sense, sense ID, and interrogate. Each command initiates a corresponding device operation.

    22. Re:No it hasn't by erikscott · · Score: 2

      I'm starting to suspect we're in violent agreement here. :-)

      I've physically, with my eyeballs, seen Linux running on some sort of z series a couple years ago. I saw AIX/370 running on some sort of box around 1990-92-ish, so I know it can be done (parenthetically, I was told it shared no code at all with AIX/6000). My entire point with virtualization is not to suggest there's a problem with the mainframe. Whether it makes sense to or not is completely beside the point.

    23. Re:No it hasn't by bws111 · · Score: 2

      Somehow I completely missed the fact that you were talking about the virtualization aspect of it. Sorry about that.

      Even so, I don't think it that big of a deal. Linux on Z has it's own drivers, for DASD, OSA, etc. The virtualization layer only needs to trap the 'start subchannel' instruction and translate the CCWs into the 'real' CCWs then do it's own start subchannel. There is no need for the virtualization layer to emulate things like the channel controllers because that is all invisible to the OS anyway. Z architecture is probably easier to virtualize (I/O wise) because so much of it is invisible to the program. Catch the 'start subchannel' and create an I/O interrupt when done and whatever happens in between is unknown.

    24. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      Yup - first thought that ran through my mind: "Oh, they're selling Z Series with crippled Firmware."

      They've been selling z Series with crippled firmware since time immemorial - when special "z Application Assist Processors" were introduced as core options, they allowed to speed up Java and XML processing using extra copies of exactly the same hardware as the general purpose CPU cores...that was locked out (using microcode or something) from running the regular z Series software. The same goes for the "Integrated Facility for Linux" core options that only ran Linux. And the "z Integrated Information Processors" for accelerating database processing and some other highly specific things. And of course, IBM will sell you more physical CPUs than you bought, and later tease you into unlocking them by paying extra money. (But I assume you already know all this, just adding it on record for the sake of completeness.)

      --
      Ezekiel 23:20
    25. Re:No it hasn't by Anonymous Coward · · Score: 0

      It seems apparent that there now exists a significant market segment that only cares about Linux serving and not legacy mainframe apps.

      Approximately 0% of which runs on S/390.

      Given that I can buy a 2TB Xeon server, or up to what, 16TB SPARC64 or POWER8 server, both much faster than any mainframe on the market, why would I want to buy a more expensive and slower S/390 based system to run my Linux-only infrastructure?

    26. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1, Interesting

      Presumably, what is meant by that is that their POWER business led to CPU architecture advancements that were strapped onto mainframe tech to power up (pun intended) the extended S/360 ISA. At least that's the thing that would make the most sense to me.

      --
      Ezekiel 23:20
    27. Re:No it hasn't by Anonymous Coward · · Score: 0

      Shut the fuck up with your lies.

      You can go on Youtube and hear it from the horses mouth that the Z architecture is not PowerPC:

      here (second half)

    28. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      Isn't there still software buried in the I/O stack that is kind of emulating a punch-card reader?

      Well, in your IBM PC compatible, there could be an SSD disk kind of emulating a 512B block device. ;)

      --
      Ezekiel 23:20
    29. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      Well, that segment can run Linux on whatever they want, especially since many large-scale Linux-based computing systems have already shifted the job of resiliency (implemented on mainframes in hardware) into a sea of distributed software services.

      --
      Ezekiel 23:20
    30. Re:No it hasn't by bws111 · · Score: 1

      Specialty processors like AAPs and ZIPs exist for one reason - saving the customer money. IBMs own software is mostly priced by workload - the more you use it the more you pay. But most ISV software is priced by the capacity of the machine. Since ISV software will not be dispatched on the specialty engines, those engines do not count towards the machine capacity, thereby lowering software costs.

      I am not sure exactly how 'IBM sells you more than you bought'. Perhaps what you meant is that under certain conditions you get more physical hardware than you paid for, but can't use it. So what?

    31. Re:No it hasn't by HiThere · · Score: 1

      Traditionally the main advantage of mainframes was expensive high-capacity peripheral devices. (Well, they used to also be a lot faster...but they were enough more expensive that this was a dubious advantage over a network of micro-computers. And, of course, depending on which decade you are looking at.)

      So. Perhaps the advantage is that this is a mainframe built totally around high end commodity CPUs. IBM sold they chip fab awhile ago, so there's not too much reason that they should stick with the power architecture.

      (OTOH, if I really cared I'd go read the original article and see if it provided more info. As it is...well, it's been several decades since I actually used a mainframe...)

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    32. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      They could equally well save the customer money by not crippling those cores, wouldn't you think? One wonders why ISV software couldn't simply be licensed so that it won't run on more cores that the license allows. Or maybe it could get actually priced by workload and you'd get billed on basis of transaction stats.

      --
      Ezekiel 23:20
    33. Re:No it hasn't by Tough+Love · · Score: 1

      Reliability.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    34. Re:No it hasn't by Tough+Love · · Score: 1

      These days, mainframe peripherals are much like any other... high end EMC or HDS gear or the like, 10G or 40G ethernet and possibly fiberchannel (dinosaur still walking the earth). Interconnect on the mainframe would be better in general than high end rack mount servers, a modest advantage but not enough to justify the cost. Availability is the trump card.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    35. Re:No it hasn't by Anonymous Coward · · Score: 0

      Well, in your IBM PC compatible, there could be an SSD disk kind of emulating a 512B block device. ;)

      Sure, there's also timer hardware buried somewhere in the motherboard chipset running at 4/3rds of the NTSC color subcarrier frequency. And even that is newer technology than punch-card readers.

    36. Re:No it hasn't by Shinobi · · Score: 1

      As to the interconnect, there's also Infiniband. However, the BIG thing with mainframe I/O is that unlike the PC hardware, there's no single system bus that gets split between devices, instead you have a crap-ton of channels that can all communicate concurrently, and the available features for data integrity, encryption etc, which includes checksumming of transfers between devices or between device and RAM etc.

    37. Re:No it hasn't by bws111 · · Score: 1

      Specialty engines are not crippled in any way, they always run at full speed. Even when the CPs are not full speed. And they cost less than full speed CPs. As for why don't the ISVs change how they license or bill, you would have to ask the ISVs. My guess is they don't want to do extra work and make their code more complicated for the sole purpose of getting paid less.

    38. Re:No it hasn't by Tough+Love · · Score: 1

      ...the BIG thing with mainframe I/O is that unlike the PC hardware, there's no single system bus that gets split between devices, instead you have a crap-ton of channels that can all communicate concurrently...

      You mean, unlike old PC hardware. Modern AMD hardware has hypertransport and Intel followed with QPI, both point to point serial architectures. The traditional shared bus is now just an emulation.

      and the available features for data integrity, encryption etc, which includes checksumming of transfers between devices or between device and RAM etc.

      PCI devices (now actually serial-connected) are unrestricted in what they can do. Network hardware has because a lot smarter and does a lot more offload. Disk hardware, not so much, but it's heading in that direction. So basically the same thing, which should be no surprise. Updated and subject to stronger evolutionary selection.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    39. Re:No it hasn't by Megol · · Score: 1

      The IBM Z mainframe is a direct descendant from the IBM 360 from the 60ies.

      Using a modified PPC to run the legacy CISCy code would be bad both for performance and reliability. Even assuming you meant IBM POWER rather than Power PC this holds.

      IBM shares process technology and experiences in e.g. optimizing decimal floating point execution between the POWER and the Z series - but they are completely separate designs.

    40. Re:No it hasn't by Shinobi · · Score: 1

      Hypertransport and QPI are only between CPU and either a communications hub or straight to the northbridge, and the internal shared bus(AMD and Intel have both shied away from Crossbar Switches, though HP has a custom one for their Superdome machines(which itself tries to go into Big Iron areas)). Things that would make my current Sandy Bridge Xeon or my previous Opteron system choke, such as trying to make multiple infiniband cards run at full steam simultaneously to and from a RAM disk didn't even register as a blip on system load on a mainframe when my software set was tested.

      As for the offload, this isn't individual devices, but rather additions to the communications hub that give you more feature, such as hardware checksumming on the actual transfer between device and RAM, or on-board hardware crypto, essentially zero load to the CPU, and independent of any devices.

    41. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      I meant "crippled" as in "purposefully blocking the execution of some kinds of software" (which is crippling in the sense that the hardware could run that software technically just fine if the mainframe vendor didn't decide on some random marketing-related reason that it won't be allowed to do that).

      --
      Ezekiel 23:20
    42. Re:No it hasn't by K.+S.+Kyosuke · · Score: 1

      From what I understand, HyperTransport can be used between whatever devices that are HyperTransport-enabled. Just because it doesn't get used to connect storage in PCs doesn't necessarily mean it couldn't plausibly be used to connect storage in a different, non-PC-compatible architecture that coincidentally settled on Opterons as its CPUs. Or did I miss something?

      --
      Ezekiel 23:20
    43. Re:No it hasn't by Tough+Love · · Score: 1

      You have it right. In particular, cpus may be connected to to other cpus by hypertransport, implementing SMP

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    44. Re:No it hasn't by Tough+Love · · Score: 1

      Time to repeat your test on a high end Broadwell. Also, check that each infiniband card is bound to a dedicated core and that your memory regions are properly NUMA-bound to the respective core. With proper configuration I expect you will be able to run multiple Infiniband cards at full speed (personally, I doubt Infiniband has a future with the industry piling on to RDMAoE, but that's another story).

      In any case, there is no fundamental difference in the engineering problem to be solved or the designs used to solve it. Just different tradeoffs for power efficiency and transistor budget, with commodity server CPUs looking more like mainframe processors all the time. Intel goes where the data center mass market is, and because the mass market is galloping in the direction of high IO, Intel will continue to lather on more IO with each generation. AMD likewise of course, to the extent that their monopoly-abused finances permit it. Massive amounts of PCI-connected flash is driving this, for one thing.

      I suspect you will find that Broadwell has already closed the IO gap to a small sliver.

      --
      When all you have is a hammer, every problem starts to look like a thumb.
    45. Re:No it hasn't by Anonymous Coward · · Score: 0

      Mainframes break like everything does.

      If you're smart, you build reliability into your system protocol. You can buy a lot of redundant nodes, redundant switches, redundant routers and redundant power supplies for the price of one, count 'em, one mainframe.

    46. Re:No it hasn't by Blaskowicz · · Score: 1

      I bet you meant Haswell-E, Broadwell-E is more of the same but in 2016 or late 2016.

    47. Re:No it hasn't by Blaskowicz · · Score: 1

      PCIe is built in the CPU, even in $40 consumer ones. I guess that's not too bad. Else have a look at next-gen Intel high-end, that goes full-on "Itanium is dead".
      One variant has what was described as a "100G" interconnect, so that's something you might use though that's yet another interconnect.

      http://www.cpu-world.com/news_...

      The Skylake-F will incorporate one link of first generation Storm Lake Fabric. The Skylake-F will use Multi-Chip packaging, that will house two dies, a Skylake die and a Storm Lake die. In all other aspects, the Skylake-F will be identical to the Skylake-EP.

    48. Re:No it hasn't by Blaskowicz · · Score: 1

      From wikipedia

      Each core has six RISC-like execution units, including two integer units, two load-store units, one binary floating point unit and one decimal floating point unit. The zEC12 chip can decode three instructions and execute seven operations in a single clock cycle.[4] Attached to each core is a special co-processor accelerator unit; in the previous z CPU there were two shared by all four cores.

      Sounds just like x86.

    49. Re: No it hasn't by Anonymous Coward · · Score: 0

      IBM can't control ISV pricing.

    50. Re:No it hasn't by Blaskowicz · · Score: 1

      I'm tired at the state of PC gaming, so I guess I'm gonna use an old Pentium II or III one of these days, with ISA slots to get full sound support under DOS. To make it neat and silent.. there's software to run DOS from iSCSI :)

    51. Re:No it hasn't by Shinobi · · Score: 1

      Hypertransport and QPI both cause problems with communication between devices on different PCI-E switches, that's one of the things Infiniband lets you get around(and Infiniband is about so much more than just RDMA, it's also about the flexible structure you can tailor for your use, and maintain single-microsecond latencies, with all the bells and whistles. Despite the hurfblurf, RDMAoE has not made that happen).

      Here's an example of the issues involved: http://www.cirrascale.com/blog...

    52. Re:No it hasn't by Shinobi · · Score: 1

      Here's an example of the issues with I/O that involves PCI-E:

      http://www.cirrascale.com/blog...

  8. So it competes with SUN. by LWATCDR · · Score: 1

    Large, fast, massive IO, and extremely reliable computers running a POSIX or POSIX like OS.

    --
    See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    1. Re:So it competes with SUN. by nhtshot · · Score: 1, Interesting

      Don't forget that Z series is even more stupid expensive than Sun gear. I get that there's a bunch of R/D that goes into mainframes and keeping a non x86 CPU alive (Sparc/PPC/Zseries). But, if you want new things to be built for them, there has to be a reasonable level of entry for small shops.

      $20k+ (Sun) or $100k+ (Zseries) is not a low enough entry level that I'd going to develop anything for it.

    2. Re:So it competes with SUN. by GCsoftware · · Score: 1

      You do realise you can emulate a z/Series using Hercules and use that for development, right?

      No need to fork out $100K+ (also since these Linux-only z boxes won't have any CP's, just zIFLs, I think the base price might come down)

      And since it's pretty easy to set up virtualised instances (either using LPARs or say z/VM), you could always share one of these boxes amongst multiple users.

  9. Really? by Anonymous Coward · · Score: 0

    Correct me if i'm wrong, but isn't the sole reason for running and mainframe either to run legacy applications/software/images _OR_ software that can guarantee certian conditions of reability/transactions/processing. I'm aware that z-Series and i-Series machines can boot up Linux images for some years now, but it feels a bit of a niché.

    Or will IBM add this "guarantee" to software like the kernel, Apache, PostgreSQL, etc.., or else it will be only a astronomical expensive Linux VPS. It's sure is fantstic machines, and Linux will probably run lovely on them. But at what gained value, and cost?

    1. Re:Really? by Anonymous Coward · · Score: 0

      é, used in places like "centré" and "niché" to reveal that the writer learned a little something in French class, but not enough.

    2. Re:Really? by Anonymous Coward · · Score: 5, Informative

      Mainframes are nice in that you get hw with 100% uptime. Not 99.99%, but 100%. Electronics getting old? Need replacements? Offline a couple of CPUs, then pull the cards while the machine is running. Insert new ones and bring them up. Repeat, until you've swapped all the CPUs - and the mainframe was running all the time! (Obviously not at 100% capacity, but transactions were processed continuously).

      Memory modules are hot-replaceable in the same manner. So is network, disks & power supplies. All is redundant, all is replaceable without shutting down. You can do such stunts "to some extent" with PC hardware - i.e. you can get a pc-type server board with redundant power. And linux has hot-adding of CPUs already. But mainframes has 50 years of experience with this sort of always-up requirement - so it just works, without snags.

    3. Re:Really? by Second_Derivative · · Score: 1

      How do you replace the motherboard? What if a cap blows?

      Something has to hold the system bus, and that can't be hot-plugged.

    4. Re:Really? by GCsoftware · · Score: 4, Informative

      Nokia's DX200 series of PSTN switches had fully redundant motherboards, you could literally physically cut the PCI bus and the thing would just keep on rolling, without dropping any calls..

    5. Re:Really? by bws111 · · Score: 2

      The logic is packaged in 'drawers' (up to 4 per system). If one fails it is taken offline and replaced and the image keeps running (at lower capacity of course).

    6. Re:Really? by Shinobi · · Score: 2

      Mainframes don't have a system bus in the way the PC crowd thinks of it. You can in fact swap out the backplane parts one at a time and maintain system/image uptime/integrity.

    7. Re:Really? by RabidReindeer · · Score: 5, Informative

      Mainframes are not simply overpriced PCs. They're put together internally in quite a different way.

      The original system busses were in the backplane, not in on a "motherboard". That was true even on my very first (S-100) PC, long before IBM got into the personal computer market. The backplane was almost nothing but wiring, with no caps to blow. You'd basically have to set it on fire to render it useless.

    8. Re:Really? by Anonymous Coward · · Score: 1

      Since "C" generally really is "K" in Latin, but people anyway seem quite happy to talk about stuff like"Sentrums" and "Seramics", I don't quite see your point.

    9. Re:Really? by Junta · · Score: 1

      I don't know about system z, but in some designs there are even redundant midplanes/backplanes, such that you could service them independently.

      Yes 'bunch of wires' (traces) can have problems. A metal can corrode, a connector can deform. It's one reason that if you have a fully redundant system and expecting 100% uptime, sometimes a midplane is a worse decision than discretely cabled components. However with redundant and indpendently serviceable mid/backplanes, that no longer becomes a risk.

      --
      XML is like violence. If it doesn't solve the problem, use more.
    10. Re:Really? by Guy+Harris · · Score: 1

      Since "C" generally really is "K" in Latin, but people anyway seem quite happy to talk about stuff like"Sentrums" and "Seramics", I don't quite see your point.

      His point is presumably that accenting the "e" at the end of "niche" reveals that the person doing so learned about accents, but didn't learn that "niche" doesn't have an accent over the "e" in French, in French class.

    11. Re:Really? by Anonymous Coward · · Score: 0

      The logic is packaged in 'drawers' (up to 4 per system). If one fails it is taken offline and replaced and the image keeps running (at lower capacity of course).

      And all the applications running on that drawer are dead, and have to be restarted from the last checkpoint.

      This is really no different to a traditional database cluster running on PC or unix server hardware.

      The bandwidth between the drawers also sucks compared to SPARC and POWER systems, which also have this redundant drawer methodology. It's much higher bandwidth than PCs with commodity interconnect, but in reality you can achieve the same result much less expensively with a PC cluster and some Infiniband switches.

      Mainframes are for business jerks, who want to feel like they are running a Real Business (tm), so shell out more money for hardware that the IBM business jerk tells them is the only reliable computer system.

      Disclaimer: I work for business jerks who buy both Mainframe and POWER DB2 installs, and both systems are capable of pretty much the same thing, but the Mainframe option is much more expensive. I work for other guys who do the same thing with Linux PCs and Oracle RAC or DB2 or even Postgres.

      Mainframes are no different to any other redundant computing system, whether it be a competitive Unix system with partitioned CPU domains, or a rack of PCs running a redundant application stack.

    12. Re:Really? by Blaskowicz · · Score: 1

      Niché does mean something along the line of "nidificated".

  10. It should be more about COBOL? by Anonymous Coward · · Score: 0

    The issue with Mainframe is that there are still customers running MVS because they are locked with their COBOL applications. Is it COBOL good or bad? Does it makes sense making new developments in COBOL? I don't know. Plus there is the evil Monthly License Charge (MLC) on older mainframes, which means that you "rent" the software and if you stop paying you must unplug your mainframe. It is time for IBM to help customers to "really" move away from MVS and older mainframe infrastructure instead of looking at the older mainframe as "the legacy that gives me some cash".

    1. Re:It should be more about COBOL? by dimeglio · · Score: 1

      Yet the mainframe is by far the most reliable and secure environment where to run production software. The folks working on mainframe have traditionally more rigour in their processes than those running mid-range servers. At least it was the case not that long ago.

      --
      Views expressed do not necessarily reflect those of the author.
    2. Re:It should be more about COBOL? by Major+Blud · · Score: 1

      "Yet the mainframe is by far the most reliable and secure environment where to run production software"

      Maybe, but I always felt it was more like "security through obscurity". There aren't many z/OS exploits because of it's low usage footprint, not because it's anymore inherently secure than a modern UNIX/Linux.

      --
      If you post as Anonymous Coward, don't expect a reply.
    3. Re:It should be more about COBOL? by erikscott · · Score: 1

      It's not the mainframe that's so bulletproof. It's MVS. And, as you noted, an extremely risk-avoidant culture.

    4. Re:It should be more about COBOL? by Blaskowicz · · Score: 1

      So, is the point that if you're spending millions on it, then you will be more careful with the software?
      Do the programmers get to wear a white lab coat?

    5. Re:It should be more about COBOL? by RabidReindeer · · Score: 1

      That's because when you've coughed up a million or six for a mainframe that can take up to 10 minutes to re-IPL and paralyzes the entire company while it's down, the last answer to a technical problem you want to hear is "Have you tried turning it off and back on again?"

      The whole concept of "fixing" problems via Ctrl-Alt-Del is one of the worst things that ever happened to computing technology.

      Hardware used to be expensive, so companies hired expensive employees to provide expensive (but reliable) solutions.

      Then hardware got cheap, so companies looked for cheap employees to provide cheap solutions. And got them.

    6. Re:It should be more about COBOL? by RabidReindeer · · Score: 2

      So, is the point that if you're spending millions on it, then you will be more careful with the software?
      Do the programmers get to wear a white lab coat?

      Usually more like torn-up jeans and ratty t-shirts. The one who dress fancy are the ones who are least to be trusted.

    7. Re:It should be more about COBOL? by Anonymous Coward · · Score: 0

      "Yet the mainframe is by far the most reliable and secure environment where to run production software"

      Maybe, but I always felt it was more like "security through obscurity". There aren't many z/OS exploits because of it's low usage footprint, not because it's anymore inherently secure than a modern UNIX/Linux.

      The Unix security model is extremely limited 0/1 user/root user/system. Traditionally most mainframes have had much more sophisticated models. It's worth looking at what was implemented in both VMS and Multics as well as the various IBM operating systems. If properly implemented these various features should make the mainframes much more secure. Of course, simplicity has advantages too, however the difference is not simply obscurity.

  11. So what's the point for AIX? by LichtSpektren · · Score: 1

    Is it supported merely for legacy systems, or does IBM still find a niche use for AIX? Anybody care to enlighten me?

    1. Re:So what's the point for AIX? by bws111 · · Score: 1

      AIX has nothing to do with mainframes. AIX runs on POWER systems.

    2. Re:So what's the point for AIX? by Major+Blud · · Score: 1

      I know AIX is a holdover from RS/6000, but isn't the zSeries based on a highly modified POWER core?

      --
      If you post as Anonymous Coward, don't expect a reply.
    3. Re:So what's the point for AIX? by Anonymous Coward · · Score: 0

      Mainframes run zOS, a prosperity OS.

      https://en.wikipedia.org/wiki/Z/OS

    4. Re:So what's the point for AIX? by bws111 · · Score: 1

      No, not at all.

    5. Re:So what's the point for AIX? by Major+Blud · · Score: 1

      Looks like you're correct, although z/architecture does share some of the same components (in the same way all CPU's share some of the same technologies).

      http://speleotrove.com/decimal...

      It makes sense considering POWER is RISC whereas z/Series is CISC. I should have been able to put two-in-two together.

      --
      If you post as Anonymous Coward, don't expect a reply.
    6. Re:So what's the point for AIX? by Anonymous Coward · · Score: 0

      IBM has a software layer called UNIX System Services to provide an AIX environment on most of their Z-mainframes - apparently it won't be supported on this new one though. From my limited experience, I think the Linux implementation was more seamless and better regarded than USS, which had sort of a '90s-compatibility layer feel.

    7. Re:So what's the point for AIX? by RabidReindeer · · Score: 2

      IBM mainframes were commonly virtual machines. Unlike their predecessors, which had their instructions hard-wired into them, the System/360 and later boxes usually had some sort of "Initial MicroProgram Load" phase that kitted out the machine's NVRAM with the microcode that made them all run the common S/360 instruction set, regardless of underlying hardware, which could be quite radically different, depending on the make and model. Not unlike what Project Hercules provides, but on a much dumber level. In fact, the original floppy disk systems were used as IMPL storage devices on some machines.

    8. Re:So what's the point for AIX? by RabidReindeer · · Score: 2

      Mainframes run zOS, a prosperity OS.

      https://en.wikipedia.org/wiki/Z/OS

      Considering their license fees, it certainly should be a prosperity OS. But I think you meant "proprietary".

      Actually, mainframes have run many different OS's, some proprietary, some custom, a few open-source. Although up until about 1986, the source code for most IBM OS's was freely available.

      Mainframes these days are most likely to run zOS, zVM and/or zVSE. These are the primary IBM licensed OS products. But as I said, other OS's have been implemented as well. I think a number of universities and military installations did so at one time or another.

    9. Re:So what's the point for AIX? by Guy+Harris · · Score: 1

      IBM has a software layer called UNIX System Services to provide an AIX environment on most of their Z-mainframes - apparently it won't be supported on this new one though. From my limited experience, I think the Linux implementation was more seamless and better regarded than USS, which had sort of a '90s-compatibility layer feel.

      Because it is a compatibility layer atop OS/VS2 Multiple Virtual Storage. a/k/a MVS, or, as it's called these days with its 64-bitification, z/OS - complete with EBCDIC being the native character set (so watch out for those UN\*X programs that assume 'a' through 'z' or 'A' through 'Z' are contiguous!). If the new machines don't support z/OS, they won't support USS, either.

    10. Re:So what's the point for AIX? by Guy+Harris · · Score: 1

      IBM mainframes were commonly virtual machines. Unlike their predecessors, which had their instructions hard-wired into them, the System/360 and later boxes usually had some sort of "Initial MicroProgram Load" phase that kitted out the machine's NVRAM with the microcode that made them all run the common S/360 instruction set, regardless of underlying hardware, which could be quite radically different, depending on the make and model.

      For System/360, only the Model 85 and Model 25 had microcode in RAM rather than ROM (the Model 75 and Model 91 didn't have any microcode, the instruction set was implemented in hardwired logic). The hardware was, as far as I know, primarily designed to implement the System/3x0 instruction set; different machines may have been friendly towards other instruction sets to different degrees (emulators existed for some 140x and 709x machines, but they only needed to run those instruction sets as well as the original machines, not as well as the S/3x0 instruction set ran). Writable control store became common in System/370.

      These days, I think the commonly-used instructions are hardwired (and multi-operation instructions cracked into micro-ops, Pentium-Pro-and-successors-style, in the latest chips), and some of the more-complicated instructions may trap to "millicode", which is native machine code, perhaps with special access to machine-specific hardware. (Yes, it does sound a bit like PALcode.)

  12. And what might this cost? by GGardner · · Score: 3, Interesting

    I notice TFA has no mention of what the hardware will cost, or what IBM will charge for Linux on a mainframe, or even the model numbers of these two mainframes which are Linux only. And MongoDB on an IBM mainframe? Talk about a culture clash.

    1. Re:And what might this cost? by Anonymous Coward · · Score: 1

      If you have to ask, you can't afford it.

    2. Re:And what might this cost? by pauljuno · · Score: 1

      Oddly enough, I just read an article on Ars that mentions MongoDB specifically. http://arstechnica.com/informa...

    3. Re: And what might this cost? by cyber-vandal · · Score: 1

      Is it mainframescale?

  13. On AS/400 midranges NT's been there since 1998 by Anonymous Coward · · Score: 0

    Question:"... which IBM server range can run applications written for Windows NT and 2000, Novell NetWare, Aix and OS/2 as well as..."

    Answer: "Probably the most important development, however, came in 1998, when the ability to run Windows NT was added (Windows 2000 has become an option now on the latest version)"

    * BOTH quotes are from -> http://www.computerweekly.com/...

    APK

    P.S.=> Still, it's pretty cool seeing Linux being helped along thus by IBM (one of their BIGGEST proponents from the commercial world imo, & afaik, one of the BIGGEST contributors of helping Linux's codebase become more solid as well)... apk

    1. Re:On AS/400 midranges NT's been there since 1998 by Guy+Harris · · Score: 1

      Question:"... which IBM server range can run applications written for Windows NT and 2000, Novell NetWare, Aix and OS/2 as well as..."

      Answer: "Probably the most important development, however, came in 1998, when the ability to run Windows NT was added (Windows 2000 has become an option now on the latest version)"

      * BOTH quotes are from -> http://www.computerweekly.com/...

      ...which is talking about add-on x86 processors running NT (and other x86 operating systems).

  14. locked down hardware ? by Anonymous Coward · · Score: 0

    I guess if MS can do it, IBM can too. I also noticed the TechCrunch article has a link to the announcement:
    http://www.linuxfoundation.org/news-media/announcements/2015/08/linux-foundation-brings-together-industry-heavyweights-advance
    which produces "Access Denied"!! So I shortened it:
    http://www.linuxfoundation.org/news-media/announcements/2015/08/
    And that page shows no such announcement. WTF? Smells fishy to me.

    1. Re:locked down hardware ? by Guy+Harris · · Score: 1

      I guess if MS can do it, IBM can too.

      Except this is probably IBM locking out their own operating systems, i.e. they're not "machines that can't run anything other than Linux", they're "machines that can't run z/OS or z/VSE", which IBM has already had for a while. Given that I don't think anybody's has completed a port of any other open-source OSes to z/Architecture, that may amount to "machines that can't run anything other than Linux", but, unless there are bits of z/Architecture Linux that are binary-only and that support undocumented parts of the system (which I think there might be), that's not inherent to those systems.

      (This is more like Apple releasing a Mac that doesn't have the right magic to have OS X willing to boot on it; it could still run Linux or Windows or....)

      I also noticed the TechCrunch article has a link to the announcement: http://www.linuxfoundation.org/news-media/announcements/2015/08/linux-foundation-brings-together-industry-heavyweights-advance which produces "Access Denied"!!

      Works for me....

      So I shortened it: http://www.linuxfoundation.org/news-media/announcements/2015/08/ And that page shows no such announcement.

      It does now, at the top.

    2. Re:locked down hardware ? by bws111 · · Score: 1

      Right. The hardware does not care what you run on it. However, if you have only IFLs z/OS won't load. They probably do that by not providing some undocumented instruction that z/OS needs when the engine is configured as an IFL. The Linux stuff is all open, there are no binary-only bits or undocumented instructions used by Linux.

      There used to be a Z version of OpenSolaris. Don't know if it still exists or not. If it does, it would run on these machines.

      The 'Linux-only' part of this announcement is not that the MACHINES are only capable of running Linux, it is that it is now possible to have a virtualized Linux environment using ONLY Linux (before you had to have z/VM to provide the virtualization).

  15. Ubuntu Phone is real. No Red Hat Enterprise phone by raymorris · · Score: 2

    Except for, well the actual facts. Canonical does in fact put Ubuntu on phones. That's actually one of their products, Ubuntu Phone. Red Hat, on the other hand, sells Red Hat ENTERPRISE Linux. They do in fact have a different focus.

    "It's all Linux", one might say. Both do use (different) Linux kernels, just like Android does. There are also differences, such as the focus on new features vs time-tested reliability. Red Hat doesn't get the hottest new stuff the moment that upstream releases a beta. They wait until it's stable and reliable. For mainframes, you probably want stable and reliable.

  16. Ubuntu adware/spyware? by nickweller · · Score: 2

    What spyware, please include citations.

    Scale out with Ubuntu Server

    @Anon: "I'm sorry, but Ubuntu on mainframes? Ubuntu is the linux distribution FURTHEST from being appropriate for a mainframe - it's heavily targeted towards desktop users, particularly those with a lower level of expertise (or a lower desire to put work into their OS) than the average linux user. What's more, it's adware/spyware now, which is definitely something I'd hate to have on a mainframe - the last thing you want is your OS transmitting and receiving data at random!" ref

    1. Re:Ubuntu adware/spyware? by danomac · · Score: 1

      It does happen, although I'm not sure it happens with Ubuntu server. Canonical does allow you to turn it off, but I have no idea how hard it is to do so.

      Citation

  17. Seconded by Anonymous Coward · · Score: 0

    I'm also interested in picking one of these up. Can somebody give us a ball-park figure on this bad boy?

    1. Re:Seconded by GGardner · · Score: 1

      In 2009, the Register quoted z systems Linux mainframe pricing at $323,204 for a machine with five cpu cores, and no memory or disk: http://www.theregister.co.uk/2...

  18. Mainframe runs on Linux by tomhath · · Score: 1

    mainframe servers that only run on Linux.

    Something about that quote seems backwards to me. Can I run that server on a raspberry pi running Linux?

    1. Re:Mainframe runs on Linux by wonkey_monkey · · Score: 1

      There used to be a Windows ad along the lines of "the software that runs forfty percent of the world's computers." They later changed it, either to "runs on" or "is run by" or something like that.

      --
      systemd is Roko's Basilisk.
    2. Re:Mainframe runs on Linux by Guy+Harris · · Score: 1

      mainframe servers that only run on Linux.

      Something about that quote seems backwards to me. Can I run that server on a raspberry pi running Linux?

      Will Hercules run on an Raspberry Pi? If so, then, yes, you can run that server on a Raspberry Pi running Linux.

      (But, yes, it should have said "that only run Linux".)

    3. Re:Mainframe runs on Linux by bws111 · · Score: 1

      People are misreading the announcement (and, yes, a lot of it has to do with bad reporting). The annoucement is not a MACHINE that is only capable of running Linux. The announcement is about SOFTWARE, namely that you can now have a virtualized mainframe (ie 'cloud') environment using ONLY Linux. Previously the virtualization had to be provided with z/VM. Now it can be provided by Linux running KVM.

  19. Re:Ubuntu Phone is real. No Red Hat Enterprise pho by Blaskowicz · · Score: 1

    I tried to make as much sense as the parent. Know what? Ubuntu has been quite similar to debian especially if you consider that LTS is the real version (that's semi-official since the other ones got reduced to 9 monthes)
    Currently Ubuntu LTS is more conservative than debian jessie, since the latter has systemd. I'm not up to speed about what petty things people can troll about with command-line Ubuntu LTS though.

  20. It's Kind of Like That One Time by Greyfox · · Score: 1
    I remember back in the early '90's, IBM made a number of statements that they were putting all their chips behind OS/2, and that it was going to be OS/2 everywhere. The shiny new OS/2 they were working on for the PowerPC (Mach kernel) could boot and stay up for 20-30 seconds and they were confident that was going to improve. They were going to put OS/2 on their mid-range devices! They were going to put it on their mainframes! If you were their customer you'd only have to learn one OS and use it on everything from your ATM machine to your big iron. Oh by the way, they'd also just bought Lotus for $4 Billion, were mandating that all employees were going to use Lotus Notes for E-Mail and they were planning to port their ticketing system, RETAIN, to Lotus Notes.

    AFAIK, they never did actually kill RETAIN off. Nothing could outperform that thing, much less fucking Lotus Notes. Also AFAIK, Lotus Notes continues to suck for E-Mail inside IBM The old mainframe-based E-Mail, Profs, was a significantly better E-Mail system. And they killed OS/2 off in '95 or '96. Quite possibly the best thing about Linux everywhere on the company's hardware is that IBM Didn't Invent It.

    --

    I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

    1. Re:It's Kind of Like That One Time by tomhath · · Score: 1

      Before that they announced PL/1, the First and soon to be the Only Programming Language. Except it flopped, and never replaced any of the other programming languages in vogue at the time.

    2. Re:It's Kind of Like That One Time by Greyfox · · Score: 1

      Oh yeah! I remember talking that in some class about great computing failures! Didn't they just put all the keywords and syntax from Fortran, Pascal and COBOL into one language? Right about the same time as AT&T was inventing C, as I recall!

      --

      I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

  21. Cheaper to simulate mainframe on AWS by Anonymous Coward · · Score: 0

    Everyone should run all applications on a simulated mainframe run on AWS! Talk about best ROI.

  22. Mainframes are slooooow..... by Anonymous Coward · · Score: 0

    The only reason IBM sells mainframes is because of lock-in effects. If customers could migrate off mainframes, they would because they are very very expensive and sloooow. You need only ten Intel Xeon cpus to match the largest IBM z13 mainframe with 24 sockets.

    1) IBM sells a few 100s mainframes each year, and they account for something like 15% of IBM total revenue (which is huge).

    2) Mainframes are sloooooow. Mainframe cpus are much much slower than a decent x86 cpu. IBM refuse to publish benchmarks comparing an Mainframe cpu to a x86 cpu, why? If the IBM z13 cpu really was the fastest cpu in the world, IBM would publish benchmark top records all over the internet. But they don't. Why? Let us draw conclusions.

    First of all, you can emulate a mainframe on a laptop with the TurboHercules emulator:
    https://blogs.oracle.com/jsavit/entry/hercules_goes_commercial
    An old 8-socket Nehalem x86 server gives 3.200 MIPS under emulation, which is a decent midsized mainframe.
    https://en.wikipedia.org/wiki/Hercules_%28emulator%29#Performance

    Now consider that software emulating a different architecture incurs a 5-10x performance penalty, meaning that this 8-socket x86 server would give (5-10) x 3.200 MIPS = 16.000 - 32.000 MIPS if someone ported the mainframe software to x86. This means each old Nehalem cpu gives (16.000 - 32.000) / 8 = 2.000 - 4.000 MIPS.

    The newest and largest configured z13 mainframe gives 110.000 MIPS. If it is similar to the predecessor z12 mainframe, it has also 24 sockets. This means each z13 cpu gives 110.000 / 24 = 4.500 MIPS. This is only 50% faster than the old Nehalem cpu.

    The latest Intel Xeon E7v3 cpu is many times faster than the 8-core Nehalem cpu.
    http://www.anandtech.com/show/9193/the-xeon-e78800-v3-review/9
    E7v3 has 18 very fast cores. To be conservative, say one E7v3 core is only 50% faster than a Nehalem EX core (maybe it is 2-3x faster in reality). This means that the E7v3 cpu is 3.375x faster than the Nehalem EX cpu. So, the E7v3 gives 3.375 x (2.000 - 4.000) = 6.750 - 13.500 MIPS. Let us average and say it gives 10.125 MIPS. This is a lot faster than the slow IBM z13 cpu which gives 4.500 MIPS.

    Thus, you need only 110.000 MIPS / 10.125 = 10.86 Intel Xeon E7v3 cpus to fully match a 24 cpu IBM z13 mainframe. Let us round it to 10 Intel Xeon cpus.

    .

    .

    If you instead want to compare MHz and MIPS, a rule of thumb says that 1 MIPS == 4 MHz x86.
    http://www.mail-archive.com/linux-390@vm.marist.edu/msg18587.html

    So the 110.000 MIPS equals 440.000 MHz = 440 GHz. Each E7v3 cpu runs at 2.8 GHz. All these 18 cores gives in total 50.4 GHz. This means you need 440 GHz / 50.4 GHz = 8.73 Intel Xeon E7v3 cpus to match the largest 24 socket IBM z13 Mainframe. So we see again that you need roughly 10 Intel Xeon E7 cpus to match the largest z13 Mainframe which has 24 cpus.

    No matter how you count, you need ~10ish Intel Xeon E7v3 cpus to match the largest IBM z13 Mainframe which has 24 cpus. Ergo, the Mainframe cpus are dog slow. Much slower than a decent x86 cpu.

    1. Re:Mainframes are slooooow..... by Anonymous Coward · · Score: 0

      IBM has shut down TurboHercules because IBM is afraid that an emulated Mainframe on a x86 server, can replace a real Mainframe. If TurboHercules not where a real threat (could replace a real Mainframe in production), then IBM would not care about TurboHercules. IBM promised to release 511 patens to the open source community, but when IBM found out that TurboHercules use one of these patents, IBM attacked TurboHercules:
      http://arstechnica.com/information-technology/2010/04/ibm-breaks-oss-patent-promise-targets-mainframe-emulator/

    2. Re:Mainframes are slooooow..... by Anonymous Coward · · Score: 0

      Oh my, where to start ... maybe I shouldn't. You just don't get mainframes.

      For a start, the z13 CPUs are clocked at over 5Ghz, they certainly are not slow.

      Also, *everything* in a mainframe is at least duplicated, a single failed component will not stop a process from running. Would you trust those 10x Xeons to run $billions of transactions 24x7x365xn. A mainframe with a failing component will call home and and a technician from IBM will turn up with a spare before you even know you've got a problem.

      An Instruction in a mainframe is not a simple register add, it could be a very complex calculation that would take hundreds of clock cycles to execute on a Xeon.

      So, basically yes, if you want speed, use Xeons, if you want something that can eat and process data 24x7x365 for years on end without fail, then a mainframe is the right tool for the job

  23. The SUN has set. by Anonymous Coward · · Score: 0

    When a company open sources , it means no-one buys them anymore... I wonder if oracle is also interested to snatch another dead horse