4 ms·

Disclosure: I work at IBM. Mainframes are not my area, but what I've learned makes me salivate. The next time I launch a startup, one of these Z systems is goi

by abtinf 9y ago

Disclosure: I work at IBM. Mainframes are not my area, but what I've learned makes me salivate. The next time I launch a startup, one of these Z systems is going to be on the CapEx plan. The lowest end tier of these systems is something that could be easily afforded post series-A, and the stunning computational power is almost the least important benefit. These systems provide incredible redundancy, reliability, fault tolerance, and transaction processing speed. Decades of engineering experience have gone into building these. You could hire engineers to figure out to make systems run reliably on a cloud provider, but unless you are one of a tiny handful of unicorns that specialize in this, you are going to do a second-rate job of it. And hiring the staff with the skill set to do that is incredibly expensive and doesn't create differentiating value.

snowcrshd 9y ago
I always hear people say how reliable mainframes are. But if I have a startup that needs to distribute its application among different regions of the world, how does a mainframe help me with that? It is still a single point of failure. Has there been any sort of study, comparing commodity hardware and mainframes in this setting? I'll not argue a single mainframe is more reliable than a commodity server. It should be, given how much they cost.
- sumoboy 9y ago
  Also factor in the operational costs with running your own data center and people to keep things running 24x7. Floor space, storage, power, disaster recovery, people = $$$. Companies have moved away from mainframes and the long term costs by migrating to high powered blade servers, vmware, cobol emulation software. I'm sure these servers still find homes but the computing choices available today make boxes hard to sell IMO.
- abtinf 9y ago
  You can deploy systems to multiple locations in a variety of configurations to handle this need. Mainframes are not my are of specialty, but it looks like this PDF, "IBM z/OS Multi-Site Business Continuity," has a lot of details: https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/WP101635 https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebInde...
- krylon 9y ago
  I am not sure if this true for current hardware, but once upon a time the mainframe CPUs contained duplicates of each functional unit that ran in lock step; after every computational instruction, the results of the units are compared; if they differ, the CPU repeats the instruction once. If the results still differ, it "calls in sick" to the operating system. The OS then brings a spare CPU online and transfers the program that was running on the failing CPU to the new CPU, takes the failing one offline, and, depending on your service contract with IBM, calls home for a replacement CPU. The program does not even notice something went wrong. The next day an IBM service technician rings your data center's doorbell and replaces the faulty CPU, all without taking the machine offline. That kind of resiliency and redundancy runs throughout every aspect of the system's design. If you can afford it, having a mainframe be your single point of failure is not too bad.
  - snowcrshd 9y ago
    Thank you, this is the kind of information I was looking for! I still would like to see some real world comparisons, though. It sure sounds pretty impressive from your description!
  - Rjevski 9y ago
    This is definitely impressive, but most of it could be done with commodity hardware and open source software. Let's say I have a task queue with objects like (1, 2) and my task is to add them together and push the result in a new task queue - what's preventing me from doing that twice to make sure the end result is correct? Sure, I have to implement it manually, but at least my screen will no longer covered in vomit because of all the buzzwords.
    - krylon 9y ago
      My point was that a mainframe could be a single point of failure, but it engineered to such a degree of reliability and availability, that this is unlikely to be much of a problem. Plus, you can cluster them, you can even build clusters spanning several data centers. From a technology standpoint, I think they are amazing machines. The problem, of course, is that they are ridiculously expensive, which is why "open systems" (often meaning PCs running Windows or Linux, but also proprietary Unix system) have replaced mainframes in many places. In a way, I think it is fair to say that Google did something akin to what you propose. (I am told, that proprietary software for mainframes is actually so ridiculously expensive that the already-ridiculous cost of the hardware is not that much of a concern, actually.) As for the buzzwords, if you find a place where you are safe from them, send me post card, I might move there, too. ;-)
jclulow 9y ago
Frankly, when was the _last_ time you launched a startup? Why would you waste money on a technology stack which is essentially the physical manifestation of vendor lock-in and rent seeking, all but guaranteed to grow in cost in lock step with the size of your wallet? It's hard enough to find superlative software engineers and operations staff to work on more mainstream technologies; where are you going to recruit staff proficient in a technology stack alien to most everybody, who also want to shoulder the risk of joining your startup venture? No matter how much IBM wishes it were not so, we've reached a point where you absolutely _can_ (and should) do your computing on a substantially more open platform -- where more than one vendor can provide each layer of the stack that you build on!
- bitwize 9y ago
  You know how Microsoft is playing nice with Linux and open source these days? IBM did it first, because they knew vendor lock-in was a loser. It may continue to help them squeeze money out of existing customers, but it won't get them many new ones. So they had to pivot on their mainframe strategy. These days, IBM mainframes can not only run Linux and an open source stack on an LPAR, there's even a model of mainframe that is configured to only run Linux. You get the throughput benefits of the mainframe without having to commit to something that's not movable off the platform.
- Aloha 9y ago
  If you're running Linux on System Z there is no lock in - but in raw performance, even low end system z hardware will (generally) out perform x86 class hardware.
- abtinf 9y ago
  > when was the _last_ time you launched a startup? 6 years ago. After that, I was an early (< 10 headcount) engineering hire at a couple companies. > all but guaranteed to grow in cost in lock step with the size of your wallet? Check your premises. My general understanding - again, mainframes are not my area - is that IBM almost guarantees falling per-transaction costs over time with Z. > a technology stack alien to most everybody Let IBM maintain the mainframe and let your engineers build on top with whatever tech stack or platform they are comfortable with. IBM invented virtualization in the 60s. And there are quite a few more mainframe engineers out there than you might think.

by abtinf 9y ago

snowcrshd 9y ago

I always hear people say how reliable mainframes are. But if I have a startup that needs to distribute its application among different regions of the world, how does a mainframe help me with that? It is still a single point of failure. Has there been any sort of study, comparing commodity hardware and mainframes in this setting? I'll not argue a single mainframe is more reliable than a commodity server. It should be, given how much they cost.

sumoboy 9y ago
Also factor in the operational costs with running your own data center and people to keep things running 24x7. Floor space, storage, power, disaster recovery, people = $$$. Companies have moved away from mainframes and the long term costs by migrating to high powered blade servers, vmware, cobol emulation software. I'm sure these servers still find homes but the computing choices available today make boxes hard to sell IMO.
abtinf 9y ago
You can deploy systems to multiple locations in a variety of configurations to handle this need. Mainframes are not my are of specialty, but it looks like this PDF, "IBM z/OS Multi-Site Business Continuity," has a lot of details: https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/WP101635 https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebInde...
krylon 9y ago
I am not sure if this true for current hardware, but once upon a time the mainframe CPUs contained duplicates of each functional unit that ran in lock step; after every computational instruction, the results of the units are compared; if they differ, the CPU repeats the instruction once. If the results still differ, it "calls in sick" to the operating system. The OS then brings a spare CPU online and transfers the program that was running on the failing CPU to the new CPU, takes the failing one offline, and, depending on your service contract with IBM, calls home for a replacement CPU. The program does not even notice something went wrong. The next day an IBM service technician rings your data center's doorbell and replaces the faulty CPU, all without taking the machine offline. That kind of resiliency and redundancy runs throughout every aspect of the system's design. If you can afford it, having a mainframe be your single point of failure is not too bad.
- snowcrshd 9y ago
  Thank you, this is the kind of information I was looking for! I still would like to see some real world comparisons, though. It sure sounds pretty impressive from your description!
- Rjevski 9y ago
  This is definitely impressive, but most of it could be done with commodity hardware and open source software. Let's say I have a task queue with objects like (1, 2) and my task is to add them together and push the result in a new task queue - what's preventing me from doing that twice to make sure the end result is correct? Sure, I have to implement it manually, but at least my screen will no longer covered in vomit because of all the buzzwords.
  - krylon 9y ago
    My point was that a mainframe could be a single point of failure, but it engineered to such a degree of reliability and availability, that this is unlikely to be much of a problem. Plus, you can cluster them, you can even build clusters spanning several data centers. From a technology standpoint, I think they are amazing machines. The problem, of course, is that they are ridiculously expensive, which is why "open systems" (often meaning PCs running Windows or Linux, but also proprietary Unix system) have replaced mainframes in many places. In a way, I think it is fair to say that Google did something akin to what you propose. (I am told, that proprietary software for mainframes is actually so ridiculously expensive that the already-ridiculous cost of the hardware is not that much of a concern, actually.) As for the buzzwords, if you find a place where you are safe from them, send me post card, I might move there, too. ;-)

jclulow 9y ago

Frankly, when was the _last_ time you launched a startup? Why would you waste money on a technology stack which is essentially the physical manifestation of vendor lock-in and rent seeking, all but guaranteed to grow in cost in lock step with the size of your wallet? It's hard enough to find superlative software engineers and operations staff to work on more mainstream technologies; where are you going to recruit staff proficient in a technology stack alien to most everybody, who also want to shoulder the risk of joining your startup venture? No matter how much IBM wishes it were not so, we've reached a point where you absolutely _can_ (and should) do your computing on a substantially more open platform -- where more than one vendor can provide each layer of the stack that you build on!

bitwize 9y ago
You know how Microsoft is playing nice with Linux and open source these days? IBM did it first, because they knew vendor lock-in was a loser. It may continue to help them squeeze money out of existing customers, but it won't get them many new ones. So they had to pivot on their mainframe strategy. These days, IBM mainframes can not only run Linux and an open source stack on an LPAR, there's even a model of mainframe that is configured to only run Linux. You get the throughput benefits of the mainframe without having to commit to something that's not movable off the platform.
Aloha 9y ago
If you're running Linux on System Z there is no lock in - but in raw performance, even low end system z hardware will (generally) out perform x86 class hardware.
abtinf 9y ago
> when was the _last_ time you launched a startup? 6 years ago. After that, I was an early (< 10 headcount) engineering hire at a couple companies. > all but guaranteed to grow in cost in lock step with the size of your wallet? Check your premises. My general understanding - again, mainframes are not my area - is that IBM almost guarantees falling per-transaction costs over time with Z. > a technology stack alien to most everybody Let IBM maintain the mainframe and let your engineers build on top with whatever tech stack or platform they are comfortable with. IBM invented virtualization in the 60s. And there are quite a few more mainframe engineers out there than you might think.