RAS, Reliability, Availability, Serviceability. MTBF measured in years, Massive I/O capabilities. Integrated cryptographic hardware that can be configured for SSL acceleration or for general purpose cryptographic services.
Software
zVM - Forty years of virtualization behind the current release. It can virtualize zOS, zVSE, zLinux and will soon support Open Solaris. I have heard that the record for zLinux images virtualized under a single zVM instance is on the order of 64,000.
zOS - With roots going back to OS/360 (1960's) it is the premier operating system for most major corporations. With zOS 1.10 (available next month) it can utilize up to 64 processors with 1 TB of memory on a z10 CEC.
Now add the concept of sysplex into the mix. A sysplex can support up to 32 instances of zOS sharing a common workload. Need to shut down a system for hardware or software maintenance, the rest of the environment just keeps on running.
Now add coupling facilities into the environment. Coupling facilities are mainframes running a unique operating system that facilitates cross system communication between each system in the sysplex. A sysplex with coupling facilities is called a parallel sysplex.
Some of the features in zOS are;
The BCP provides the essential operating system services of z/OS. It includes the I/O Control Program and the z/OS UNIX System Services Kernel. z/OS XML System Services is added in z/OS V1R8.
Common Information Model (CIM) is a standard data model for describing and accessing systems management data in heterogeneous environments. It allows system administrators to write applications that measure system resources in a network with different operating systems and hardware.
Communications Server supports secure TCP/IP, SNA, and UNIX networking throughout an enterprise.
Communications Server Security Level 3 - This feature provides authentication and security services in an IP network environment. It provides support for packet filtering, tunnels, and network address translation (NAT) which enables secure communication over private and public networks. It uses the DES algorithm and it includes SSL triple DES (TDES), SNMPv3 56-bit, and IPSec TDES.
Cryptographic Services provides the following base cryptographic functions: data secrecy, data integrity, personal identification, digital signatures, and management of cryptographic keys.
Distributed File Service support provides SMB file and print serving support for Windows clients.
FFST provides immediate notification and first failure data capture for software events.
The IBM HTTP Server provides for scaleable, high performance Web serving for critical e-business applications.
IBM TDS for z/OS, introduced as a new base element in z/OS V1R8, provides client access to an LDAP server It consists of a new, rewritten LDAP server, an LDAP client, and utilities. The LDAP client and utilities can be used with the Integrated Security Services LDAP Server or the new ITDS for z/OS LDAP server.
Network File System acts as a file server to workstations, personal computers, or other authorized systems in a TCP/IP network.
z/OS Security Level 3 includes the following: OCSF Security Level 3, System SSL Security Level 3, Network Authentication Service Security Level 3, and ITDS for z/OS Security Level 3.
z/OS UNIX System Services provides the standard command interface to familiar and interactive UNIX users.
It is not your fathers operating systems any more.
z10 is a higher cost item than commodity hardware. There is no question about it, but you get what you pay for. Approximately 100% uptime, massive I/O support, concurrent hardware upgrades, concurrent microcode updates. World class engineering backed by a major hardware vendor that you might have heard of.
Since I never mentioned zOS I won't discuss pricing but you are correct zVM does have a cost involved to it. But once again you get what you pay for. Over 40 years of virtualization development support YOUR vital business.
z/VM V5.3 is designed to offer:
* Improved scalability and constraint relief
o Support for more than 128 GB real storage
o Up to 32 real processors in a single z/VM image
o Enhanced memory management for Linux guest
o Enhanced memory utilization using VMRM between z/VM and Linux guests
o HyperPAV support for IBM System Storage DS8000
o Enhanced FlashCopy support
* Virtualization technology and Linux enablement
o Support for IBM System z specialty engines (processors)
o Enhanced VSWITCH and guest LAN usability
o Modified Indirect Data Address Words (MIDAWs) for guests
o Guest ASCII console support
o Enhanced SCSI support
* Network virtualization
o Improved virtual network management
o Enhanced failover support for IPv4 and IPv6 devices
o Virtual IP Address (VIPA) support for IPv6
* Security
o Delivery of LDAP server and client
o Enhanced system security with longer passwords
o Conformance with industry standards
o SSL server enhancements
o Tape data protection with support for encryption
* Systems management
o Enhanced management functions for Linux and other virtual images
o New function level for DirMaint
o Enhancements to the Performance Toolkit
o Enhanced guest configuration
Linux under zVM is available for FREE, but you will pay for support, just like everywhere else. Do you think that Google doesn't have a support contract somewhere. As far as their customized software goes it would have been easier and cheaper to write if the underlying hardware and software was as reliable as a z10 and zVM.
And do what work? The 800,000 Linux servers I mentioned would have 4-8GB of virtualized storage and be able to run Oracle or DB2 databases, webserving, or any other production work you could throw at it.
One of the reasons why Google has to run clustering software is to fight against the expected hardware failures. A modern mainframe with modern DASD devices simply doesn't fail in such a way that takes stuff down.
In my 13 years at my current position we have had one hardware failure that took a box down. IBM was as upset as we were at the failure, diagnosed what the problem was and fixed it not only for us for for everyone else so it would never occur again.
A handful of fully loaded z10's could easily support Google's entire computing requirements.
You are correct, a z10 with all the other "stuff" is a lot more (not even in the same ballpark) as racks of massive servers.
Some of the things you do gain however are the ability to create logical partitions (max of 64) on a z10. Each partition can share the CPU's with the the hipervisor dispatching physical CPU's to each partition as necessary to dispatch work effectively.
Each partition can run zVM, virtualization software with 40 years development behind it. Each zVM instance can support the above mentioned Linux guest machines by the hundreds or thousands. zVM can easily run at 100% cpu busy and drive all the guests without any performance issues. You can't run normal servers at 100% and expect good results.
Need another guest, run a script under zVM and boot it up. Total time to create a new guest is measured in the minutes. Total cost for each new guest is close to zero.
I can fit a z10 with enough DASD to make it interesting in my 12x12 home office (might be a little tight though) I couldn't fit the google server farm in my house.
Yes just looking at the initial numbers a z10 doesn't make financial sense buy once the total cost of ownership is worked out, server administrators, power/cooling requirements, software license fees a z10 can be very competitive.
First let me state that I'm a mainframe systems programmer and a true believer of this technology. IMHO Google should start looking at mainframe based virtualization instead of the server farms they currently depend on.
One z10 complex with 64 CPU's, 1.5 TB of memory, can support thousands of Linux instances all communicating with each other using hypersocket technology. Hypersockets uses microcode to enable communications between environments without going to the actual network.
A z10 processor complex is as close to 100% fault tolerant as possible, energy efficient, cost effective when compared to the total cost of the alternatives.
I agree with you 100% percent. If management can spend 99 cents to implement a vastly inferior solution rather than one dollar to implement the perfect solution, guess which one wins every time.
The corporation that I work for experienced a similar accident several years ago. We now use a hardware based solution to encrypt all tape data that leaves the data center.
It's not too hard or expensive to do, all it takes is the will to do it.
I guarantee that the required "management will" will be enhanced once they are threatened with jail time, or the loss of their multi-million dollar pay checks for any future data loss of this type.
One of the original IBM System S360 programs, IEFBR14 is still in wide use today.
IEFBR14 CSECT
SR 15,15
BR 14
END
Only two changes in over 40 years. It doesn't do much, in fact nothing except set a zero return code, but it is widely used for dataset allocation purposes in batch dataset allocation processing.
Really? I have a FreeBSD file/print/mp3 server. I use OpenBSD running on a Soekris 4801 as my home firewall/gateway and my main home workstation is also running FreeBSD. I almost forgot, I use a FreeBSD workstation at my office that serves me well so I can perform my system programming job duties.
I don't think I will be shutting everything down today:-).
And the punch card was based, rumor has it, on the size of the US Dollar bill when Herman Hollerith http://en.wikipedia.org/wiki/Herman_Hollerith/ developed a mechanical tabulator in order to rapidly tabulate statistics from millions of pieces of data.
How does the printout get from machine one to machine two without a manual effort? I wouldn't want them to even touch each other. I feel that it is important for the voter to be sure that their vote is validated and secure.
I suggested bar codes since they are very easy for a computer to both print and read. They are also harder to forge than a simple text based ballot would be.
Several weeks ago I sent a letter to the editor to my local newspaper (which never was published) that described a much better (in my opinion) system of voting.
1) Touch screen computer that allows voter to make their choices. This machine doesn't count votes it just produces a piece of paper with bar codes printed on it. These bar codes identify all the selections the voter selected.
2) After the paper is printed the voter inserts it into a scanner computer which reads the votes and displays the selections on a screen. If the voter agrees with the display the vote is accepted. This paper then gets automatically stamped with a unique serial number and stored in a locked box.
Machine #1 and #2 must be produced by different companies so as to reduce the possibility of collusion between the printer and the scanner.
If a recount is ever needed it would be easy to put all the forms through a bar code reader and recount all the votes. It would also make sure that no forms were lost since any missing serial numbers would be obvious and it would also make it near impossible to stuff the ballot box for the same serial number reason.
There shouldn't be any privacy issues since it wouldn't be possible to tie an individual voter to a specific serial number.
I think you didn't understand what I meant to say. Each z9 computer can have up to 54 processors. The zOS 1.9 operating system can utilize all 54 processors. The parallel sysplex technology allows up to 32 operating systems to work in parallel. That works out to 32 OS's * 54 processors == 1728.
I agree that MIPS is a meaningless term but it was a number to throw out in a hope I could impress the less informed.:-)
Trust me on this, the parallel sysplex technology is somewhat different than simple shared queues. In a parallel sysplex environment an operating system component called workload manager is able to shift any work that gets started on one processor to any of the other available processors on any system in the sysplex. This shifting of work is automatic depending on that specific transactions response time goals. We can take down an operating system and the work executing there automatically gets routed to another available system without the transaction system either knowing or caring.
One of the z9 computers in the data center where I work with only has two general purpose processors and it is able to support many hundreds of concurrent processes executing with a transaction response time measured in fractions of a second.
The mainframe hardware environment is very strong but the operating system has over 40 years of enhancements beginning with the original S/360 operating systems that were announced in the mid 60's. Applications written back then in a real storage only environment (16K real storage) still happily run today in a 64bit virtual storage environment without any changes.
First let me state that I am a zOS Systems Programmer and have been working on the platform since 1979. The latest IBM mainframe, the z9 Enterprise Class processor can be configured with up to 54 CPU's with a maximum of 524 gigabytes of memory.
The next release of the zOS operating system, Version 1 Release 9, can support all 54 CPU's in a single operating system image.
With IBM's Parallel Sysplex technology we can have up to 32 operating system images working in tandem processing work in shared queues.
Do the math, 54 CPU's times 32 images. That's 1728 total processors with each one capable of over 450 MIPS. I haven't even mentioned the mainframes I/O capabilities.
Sorry, Linux on Intel is just a baby compared to mainframes.
We have an entire system that produces documents and letters that we send to to our customers. This system has to use a 1990's version of the software package. These hundreds or thousands of machine generated forms and letters were composed using this level of software that contain known software bugs and will not format correctly if they execute using a later copy of the package.
Thankfully this system is in the process of being phased out.
Microsoft has everyone brain washed into believing that whenever they release a new operating system, or even a service pack, that it is normal and expected that application programs that worked on the earlier operating system might not work on the new operating system.
In real operating systems, (here comes the flames), people should expect that their application programs work without flaw year after year. Businesses that depend on working application programs should not stand for having to change or rewrite their applications every time the operating system changes.
I work in the IBM mainframe environment (but run FreeBSD/OpenBSD at home) and I never have to worry that our applications don't work whenever we change operating systems. As long as we write to the supported system interfaces, IBM takes great pains to ensure that everything works going forward.
The mainframe (MVS) environment has had many major architectural changes in forty years but applications written back then still work. From 24 bit addressing in the 1960's, 70's and early 80's to 31 bit addressing until the advent of 64 bit addressing in the current operating systems, the old applications continue to run.
Microsoft should be ashamed of its efforts to ensure upward capability.
Slashdot Mirror
Hardware
RAS, Reliability, Availability, Serviceability. MTBF measured in years, Massive I/O capabilities. Integrated cryptographic hardware that can be configured for SSL acceleration or for general purpose cryptographic services.
Software
zVM - Forty years of virtualization behind the current release. It can virtualize zOS, zVSE, zLinux and will soon support Open Solaris. I have heard that the record for zLinux images virtualized under a single zVM instance is on the order of 64,000.
zOS - With roots going back to OS/360 (1960's) it is the premier operating system for most major corporations. With zOS 1.10 (available next month) it can utilize up to 64 processors with 1 TB of memory on a z10 CEC.
Now add the concept of sysplex into the mix. A sysplex can support up to 32 instances of zOS sharing a common workload. Need to shut down a system for hardware or software maintenance, the rest of the environment just keeps on running.
Now add coupling facilities into the environment. Coupling facilities are mainframes running a unique operating system that facilitates cross system communication between each system in the sysplex. A sysplex with coupling facilities is called a parallel sysplex.
Some of the features in zOS are;
The BCP provides the essential operating system services of z/OS. It includes the I/O Control Program and the z/OS UNIX System Services Kernel. z/OS XML System Services is added in z/OS V1R8.
Common Information Model (CIM) is a standard data model for describing and accessing systems management data in heterogeneous environments. It allows system administrators to write applications that measure system resources in a network with different operating systems and hardware.
Communications Server supports secure TCP/IP, SNA, and UNIX networking throughout an enterprise.
Communications Server Security Level 3 - This feature provides authentication and security services in an IP network environment. It provides support for packet filtering, tunnels, and network address translation (NAT) which enables secure communication over private and public networks. It uses the DES algorithm and it includes SSL triple DES (TDES), SNMPv3 56-bit, and IPSec TDES.
Cryptographic Services provides the following base cryptographic functions: data secrecy, data integrity, personal identification, digital signatures, and management of cryptographic keys.
Distributed File Service support provides SMB file and print serving support for Windows clients.
FFST provides immediate notification and first failure data capture for software events.
The IBM HTTP Server provides for scaleable, high performance Web serving for critical e-business applications.
IBM TDS for z/OS, introduced as a new base element in z/OS V1R8, provides client access to an LDAP server It consists of a new, rewritten LDAP server, an LDAP client, and utilities. The LDAP client and utilities can be used with the Integrated Security Services LDAP Server or the new ITDS for z/OS LDAP server.
Network File System acts as a file server to workstations, personal computers, or other authorized systems in a TCP/IP network.
z/OS Security Level 3 includes the following: OCSF Security Level 3, System SSL Security Level 3, Network Authentication Service Security Level 3, and ITDS for z/OS Security Level 3.
z/OS UNIX System Services provides the standard command interface to familiar and interactive UNIX users.
It is not your fathers operating systems any more.
z10 is a higher cost item than commodity hardware. There is no question about it, but you get what you pay for. Approximately 100% uptime, massive I/O support, concurrent hardware upgrades, concurrent microcode updates. World class engineering backed by a major hardware vendor that you might have heard of.
Since I never mentioned zOS I won't discuss pricing but you are correct zVM does have a cost involved to it. But once again you get what you pay for. Over 40 years of virtualization development support YOUR vital business.
z/VM V5.3 is designed to offer:
* Improved scalability and constraint relief
o Support for more than 128 GB real storage
o Up to 32 real processors in a single z/VM image
o Enhanced memory management for Linux guest
o Enhanced memory utilization using VMRM between z/VM and Linux guests
o HyperPAV support for IBM System Storage DS8000
o Enhanced FlashCopy support
* Virtualization technology and Linux enablement
o Support for IBM System z specialty engines (processors)
o Enhanced VSWITCH and guest LAN usability
o Modified Indirect Data Address Words (MIDAWs) for guests
o Guest ASCII console support
o Enhanced SCSI support
* Network virtualization
o Improved virtual network management
o Enhanced failover support for IPv4 and IPv6 devices
o Virtual IP Address (VIPA) support for IPv6
* Security
o Delivery of LDAP server and client
o Enhanced system security with longer passwords
o Conformance with industry standards
o SSL server enhancements
o Tape data protection with support for encryption
* Systems management
o Enhanced management functions for Linux and other virtual images
o New function level for DirMaint
o Enhancements to the Performance Toolkit
o Enhanced guest configuration
Linux under zVM is available for FREE, but you will pay for support, just like everywhere else. Do you think that Google doesn't have a support contract somewhere. As far as their customized software goes it would have been easier and cheaper to write if the underlying hardware and software was as reliable as a z10 and zVM.
And do what work? The 800,000 Linux servers I mentioned would have 4-8GB of virtualized storage and be able to run Oracle or DB2 databases, webserving, or any other production work you could throw at it.
Mainframes unlike consumer or even business grade servers simply don't fail in the way you described.
Well lets do some math here;
One z10, 64 lpars each running an instance of zVM. Each zVM can support a couple of thousand virtual Linux guests.
Now take five of them. 64*5*2500=800000 Linux images.
It boggles my mind and I've been an IT professional for almost thirty years.
One of the reasons why Google has to run clustering software is to fight against the expected hardware failures. A modern mainframe with modern DASD devices simply doesn't fail in such a way that takes stuff down.
In my 13 years at my current position we have had one hardware failure that took a box down. IBM was as upset as we were at the failure, diagnosed what the problem was and fixed it not only for us for for everyone else so it would never occur again.
A handful of fully loaded z10's could easily support Google's entire computing requirements.
You are correct, a z10 with all the other "stuff" is a lot more (not even in the same ballpark) as racks of massive servers.
Some of the things you do gain however are the ability to create logical partitions (max of 64) on a z10. Each partition can share the CPU's with the the hipervisor dispatching physical CPU's to each partition as necessary to dispatch work effectively.
Each partition can run zVM, virtualization software with 40 years development behind it. Each zVM instance can support the above mentioned Linux guest machines by the hundreds or thousands. zVM can easily run at 100% cpu busy and drive all the guests without any performance issues. You can't run normal servers at 100% and expect good results.
Need another guest, run a script under zVM and boot it up. Total time to create a new guest is measured in the minutes. Total cost for each new guest is close to zero.
I can fit a z10 with enough DASD to make it interesting in my 12x12 home office (might be a little tight though) I couldn't fit the google server farm in my house.
Yes just looking at the initial numbers a z10 doesn't make financial sense buy once the total cost of ownership is worked out, server administrators, power/cooling requirements, software license fees a z10 can be very competitive.
First let me state that I'm a mainframe systems programmer and a true believer of this technology. IMHO Google should start looking at mainframe based virtualization instead of the server farms they currently depend on.
One z10 complex with 64 CPU's, 1.5 TB of memory, can support thousands of Linux instances all communicating with each other using hypersocket technology. Hypersockets uses microcode to enable communications between environments without going to the actual network.
A z10 processor complex is as close to 100% fault tolerant as possible, energy efficient, cost effective when compared to the total cost of the alternatives.
I agree with you 100% percent. If management can spend 99 cents to implement a vastly inferior solution rather than one dollar to implement the perfect solution, guess which one wins every time. The corporation that I work for experienced a similar accident several years ago. We now use a hardware based solution to encrypt all tape data that leaves the data center. It's not too hard or expensive to do, all it takes is the will to do it. I guarantee that the required "management will" will be enhanced once they are threatened with jail time, or the loss of their multi-million dollar pay checks for any future data loss of this type.
One of the original IBM System S360 programs, IEFBR14 is still in wide use today. IEFBR14 CSECT SR 15,15 BR 14 END Only two changes in over 40 years. It doesn't do much, in fact nothing except set a zero return code, but it is widely used for dataset allocation purposes in batch dataset allocation processing.
Doesn't the sun also revolve around the earth also?
Really? I have a FreeBSD file/print/mp3 server. I use OpenBSD running on a Soekris 4801 as my home firewall/gateway and my main home workstation is also running FreeBSD. I almost forgot, I use a FreeBSD workstation at my office that serves me well so I can perform my system programming job duties.
:-).
I don't think I will be shutting everything down today
They also sometimes get their tops sheared off.
In Germany, they came first for the Communists, And I didn't speak up because I wasn't a Communist;
And then they came for the trade unionists, And I didn't speak up because I wasn't a trade unionist;
And then they came for the Jews, And I didn't speak up because I wasn't a Jew;
And then . . . they came for me . . . And by that time there was no one left to speak up."
Pastor Martin Niemöller (1892-1984)
As my Grandmother used to say, Good riddance to bad rubbish.
Since I now live in the Tampa Florida area and have learned about the history of the city this fits in rather well with this topic.
About 100 years ago Tampa had a huge cigar industry. The cigar workers would hire a person to sit and read the newspaper aloud to them as they worked.
Think of all the money the newspaper publishers could have received if they sued from the Cuban cigar workers back then.
It just boggles my mind.
Or bra bombers. Can you imagine the outcry from women if they had to have their bras inspected before they could fly.
I program in S/390 assembler and 80 columns are more than enough.
Column 1 - Optional Label, usual length up to 8 characters
Column 10 - Operation Code
Column 16 - Operand
Leave one space after the operand and begin comments up to column 71
Plenty of room
And the punch card was based, rumor has it, on the size of the US Dollar bill when Herman Hollerith http://en.wikipedia.org/wiki/Herman_Hollerith/ developed a mechanical tabulator in order to rapidly tabulate statistics from millions of pieces of data.
How does the printout get from machine one to machine two without a manual effort? I wouldn't want them to even touch each other. I feel that it is important for the voter to be sure that their vote is validated and secure. I suggested bar codes since they are very easy for a computer to both print and read. They are also harder to forge than a simple text based ballot would be.
Several weeks ago I sent a letter to the editor to my local newspaper (which never was published) that described a much better (in my opinion) system of voting.
1) Touch screen computer that allows voter to make their choices. This machine doesn't count votes it just produces a piece of paper with bar codes printed on it. These bar codes identify all the selections the voter selected.
2) After the paper is printed the voter inserts it into a scanner computer which reads the votes and displays the selections on a screen. If the voter agrees with the display the vote is accepted. This paper then gets automatically stamped with a unique serial number and stored in a locked box.
Machine #1 and #2 must be produced by different companies so as to reduce the possibility of collusion between the printer and the scanner.
If a recount is ever needed it would be easy to put all the forms through a bar code reader and recount all the votes. It would also make sure that no forms were lost since any missing serial numbers would be obvious and it would also make it near impossible to stuff the ballot box for the same serial number reason.
There shouldn't be any privacy issues since it wouldn't be possible to tie an individual voter to a specific serial number.
Like I said, I think that it's a good idea.
I think you didn't understand what I meant to say. Each z9 computer can have up to 54 processors. The zOS 1.9 operating system can utilize all 54 processors. The parallel sysplex technology allows up to 32 operating systems to work in parallel. That works out to 32 OS's * 54 processors == 1728.
:-)
I agree that MIPS is a meaningless term but it was a number to throw out in a hope I could impress the less informed.
Trust me on this, the parallel sysplex technology is somewhat different than simple shared queues. In a parallel sysplex environment an operating system component called workload manager is able to shift any work that gets started on one processor to any of the other available processors on any system in the sysplex. This shifting of work is automatic depending on that specific transactions response time goals. We can take down an operating system and the work executing there automatically gets routed to another available system without the transaction system either knowing or caring.
One of the z9 computers in the data center where I work with only has two general purpose processors and it is able to support many hundreds of concurrent processes executing with a transaction response time measured in fractions of a second.
The mainframe hardware environment is very strong but the operating system has over 40 years of enhancements beginning with the original S/360 operating systems that were announced in the mid 60's. Applications written back then in a real storage only environment (16K real storage) still happily run today in a 64bit virtual storage environment without any changes.
Thats amazing.
First let me state that I am a zOS Systems Programmer and have been working on the platform since 1979. The latest IBM mainframe, the z9 Enterprise Class processor can be configured with up to 54 CPU's with a maximum of 524 gigabytes of memory.
The next release of the zOS operating system, Version 1 Release 9, can support all 54 CPU's in a single operating system image.
With IBM's Parallel Sysplex technology we can have up to 32 operating system images working in tandem processing work in shared queues.
Do the math, 54 CPU's times 32 images. That's 1728 total processors with each one capable of over 450 MIPS. I haven't even mentioned the mainframes I/O capabilities.
Sorry, Linux on Intel is just a baby compared to mainframes.
We have an entire system that produces documents and letters that we send to to our customers. This system has to use a 1990's version of the software package. These hundreds or thousands of machine generated forms and letters were composed using this level of software that contain known software bugs and will not format correctly if they execute using a later copy of the package.
Thankfully this system is in the process of being phased out.
Mark Jacobs
Microsoft has everyone brain washed into believing that whenever they release a new operating system, or even a service pack, that it is normal and expected that application programs that worked on the earlier operating system might not work on the new operating system.
In real operating systems, (here comes the flames), people should expect that their application programs work without flaw year after year. Businesses that depend on working application programs should not stand for having to change or rewrite their applications every time the operating system changes.
I work in the IBM mainframe environment (but run FreeBSD/OpenBSD at home) and I never have to worry that our applications don't work whenever we change operating systems. As long as we write to the supported system interfaces, IBM takes great pains to ensure that everything works going forward.
The mainframe (MVS) environment has had many major architectural changes in forty years but applications written back then still work. From 24 bit addressing in the 1960's, 70's and early 80's to 31 bit addressing until the advent of 64 bit addressing in the current operating systems, the old applications continue to run.
Microsoft should be ashamed of its efforts to ensure upward capability.