Now, this is a very naive conclusion, stemming from utter lack of experience of how real world, or even real large systems work.
The true value of these “old” legacy systems is in years of development, debugging, validation of how these systems work. Including statistical guarantees of them working the way they were intended to. Hardware choices are thus very much secondary and dictated by how much they serve this preservation of investment in these complex systems. Rewriting them is a huge economical undertaking, not justified at all by transitioning to some new latest vogue technology buzzword, peddled by the big tech PR.
Ask HN: Why hasn't the cloud killed the mainframe?
81–88 of 88 posts
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#82Now, this is a very naive conclusion, stemming from utter lack of experience of how real world, or even real large systems work. The true value of these “old” legacy systems is in years of development, debugging, validation of how these systems work. Including statistical guarantees of them working the way they were intended to. Hardware choices are thus very much secondary and dictated by how much they serve this pr…
They also aren't exactly legacy if the performance continues to improve in such a manner that they remain impossibly far ahead of any relevant competition.
That's really the part of this whole (tiring and old) conversion that irks me the most. The total lack of willingness to come up with a better definition of legacy than "not in line with the trends that work for companies mostly interested in serving media content to consumers".
Legacy is a system that is still around that cannot be defended on its technical merits. Mainframes can and are constantly defended solely on their technical merits, and the "competition" doesn't appear to even want to try to compete there. It's always about moving the things that can be moved off platform, which is fine in the eyes of IBM, they'll even help you do it. That's why they offer much lower cost on zIIPs, zAAPs, IFLs, etc in comparison to general purpose processors running native workloads.
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#83First of all, I find I kinda funny that you call banking, retail and insurance "legacy industries". I would rather be without Netflix and Google, than banking and food ... but each to their own.. While some is inertia (mostly doing to rewriting truly large applications are hard and expensive), there is also the the point that most of those industries cannot easily handle "eventually consistent" data.. Not all transac…
I have no expertise in this area, but two counter arguments pop'ed in my head: 1: I wonder how many transactions the largest e.g Postgres clusters (or other classic RDBMS) handles per day. 500M+/day doesn't seem that incredibly high? 2: Google Spanner, which I would classify as cloudy, promises ACID guarantees at a global distributed scale. Couldn't that be used? I've listened to a Swedish developer podcast where the…
Amazon does use mainframes. They aren't a payment processor and don't handle that part of the processing. That's why the likes of VISA get to charge their fee.
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#84First of all, I find I kinda funny that you call banking, retail and insurance "legacy industries". I would rather be without Netflix and Google, than banking and food ... but each to their own.. While some is inertia (mostly doing to rewriting truly large applications are hard and expensive), there is also the the point that most of those industries cannot easily handle "eventually consistent" data.. Not all transac…
Google Cloud Bigtable and DynamoDB both appear to have ACID -- I don't see why mainframes would be better for this than cloud. Bitcoin is slow because of the many servers not in spite of it. Because of the design of the network, all servers need to receive every transaction and servers need to be able to be pretty small, which limits the transaction rate.
I heard that mainframes have redundancy built-in. /s
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#85Big tech products are decoupled from the physical world and ultimately not critical for society to function, so the workloads are highly variable. Mainframes are a good fit when you have workloads that don't vary by several orders of magnitude on a daily basis, and you need an ecosystem that treats reliability like a religion. If you haven't used z/OS or worked in a z/OS shop, you will never understand how fundamenta…
I've heard similar arguments several times, and every single time these kinds of arguments go up in flames when costs are brought up. Even discounting the "elasticity" that building applications on the cloud, its simply far cheaper from a cost perspective to build on cloud. Folks fail to mention that the cost of having the kind of all-round support from IBM that you mention comes at a significant cost that is unaffordable to most business. Its probably cheaper to debug esoteric Linux issues than to call in IBM support. In fact, IBM knows this and has moved a significant chunk of their support team abroad to reduce costs and remain affordable. In my experience, these support teams are usually of poorer quality.
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#86Earlier quoted context omitted.
Spanner does over 1B queries per SECOND (but your other point still stands of course)
If it is readonly shared-nothing queries, I am sure the hardware I have in my flat can do that as well.. The hard parts are updates in a shared system with a single consistent "view" requirement..
The hardware in your flat definitely cannot do 1 billion queries per second -- that requires a massive, global system, which can probably also support a shared system with ~100,000x less queries.
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#87First of all, I find I kinda funny that you call banking, retail and insurance "legacy industries". I would rather be without Netflix and Google, than banking and food ... but each to their own.. While some is inertia (mostly doing to rewriting truly large applications are hard and expensive), there is also the the point that most of those industries cannot easily handle "eventually consistent" data.. Not all transac…
legacy industry is sort of standard term not sure why you found it funny. It not meant to say they are less important. From quick google: "Legacy industries are those that have been around for a long time. These industries dominate a specific market and have not always had a positive approach to innovative ideas."
Legacy means highly valuable and proven. Non-legacy means unreliable and highly unlikely to ever make money. Legacy software runs the world. Without it western civilisation would collapse.
Re: Ask HN: Why hasn't the cloud killed the mainframe?
#88Big tech products are decoupled from the physical world and ultimately not critical for society to function, so the workloads are highly variable. Mainframes are a good fit when you have workloads that don't vary by several orders of magnitude on a daily basis, and you need an ecosystem that treats reliability like a religion. If you haven't used z/OS or worked in a z/OS shop, you will never understand how fundamenta…
What an excellent answer. Nailed it!