06-17-2022, 06:30 AM
@"EndtheMadnessNow" - Great contribution!
I wish you could have seen my face chuckling as I was reading through all of what you wrote! And yes, I remember all of those things, in 'painful' detail...and even more. Remember how the next generation of everything was supposed to be the end-all, be-all, for everything before it, and how many times we saw those things fail. Things like ATM (asynchronous transfer mode) , and more recently UC (unified communication)? And how many evolutions have there been in database technology, or storage, or cabling. HA!...remember "CAT-7"??? That one still cracks me up! Everyone was still marveling over CAT-5"E" from the CAT-5 world, and Anixter was running around talking about their new CAT-7 product. (and before you say it...no, CAT-7 was never even a ratified standard under IEEE, it was just some trade-name Anixter came up with to sell cable, but it "sounded" cool to the uninformed!) And I forget now what the exact bandwidth and frequencies were, but I want to say they were claiming something like 750mbps and like 450MHz. I remember cracking up at some sales dude who was trying to educate this physicist (me) about it. I asked him why I would ever want 750mbps when Cisco switches jumped from 500mbps straight to 1gbps???...who would ever benefit or use it? <crickets!> LOL!!! It was like making a "revolutionary" new fire hose which was 2.5" in diameter...in a world where there were already 1" garden hoses, and 4" fire hoses (on all fire trucks) already. It made exactly zero sense! Who was ever going to use it? Not the homeowner because their spigot was only 1", and not the Fire Dept because they already pumped 10x as much water through 4" hoses. It was one of those moments that make you go..."Hmmmmm" LOL!! But I'm digressing into the weeds of techno-jargon myself (something I hope to avoid here, at least in this thread).
<Note - To the other folks reading this...this reply is not Part III of my OP, so please don't let it distract you. This reply is my reply to the esteemed EtMN, and is a bit more technical. I promise I will get to posting Part III of 'The (Dark) "Cloud" ' here shortly, so stay tuned. So please forgive my little excursion here with EtMN. You can ignore the rest of this reply unless you're into the techno stuff>
Joking aside, given your familiarity with technology, EtMN, here's something related which is germane to the OP (albeit more technical). Remember the debate between Active-Active and Disaster Recovery (DR)? This subject is one of the major factors in why people are attracted to the "cloud". In an Active-Active environment we need extreme bandwidth, far higher than any networking QOS will allow for. In fact, there can be no QOS rules at all, it has to be 100% dedicated. Of course, this means those A-A connections are going to have to be out-of-band, right? (no "networking" allowed, except for maybe dedicated hi-bandwidth Fiber-Channel (or equal) links). Well, if an entity was at that level, then they also were already Uptime Tier III (or higher) certified and practicing physical diversity. So, to get those connections they either had to install the terrestrial links themselves, or lease the dark-fiber from the Telcos. So on the one hand they had a massive CAPEX cost, or on the other they had a massive OPEX cost, AND in order to really do A-A they had to pick one or the other. THIS is why DR stayed around as long as it did (and still does). Active-Active is incomprehensibly expensive, especially over time. So, what was the next alternative? DR, right? (now hold onto that thought for a second while I hit another point). ...
So, DR is cheaper, but it's slow (like "days" compared to "seconds", kind of slow). However, DR has an interesting selling point, and we've just seen examples and read about it all over the media recently with the catastrophic Kronos debacle, right? With DR, there is nearly zero vulnerability / risk posture because (in theory anyway) your backup is off-line. So, (again, in theory) you should be able to recover from a catastrophe (unless you've contaminated your DR backup before you realized the problem). How long has Kronos been down? 4 months? How many millions of users are there (probably 250m or more)? But you see, a system like that can't be DR; it has to be A-A because of the nature of what it does (timekeeping and payroll, fundamentally). Therein lies a paradox.
Enter, the "Cloud".
Now the host --> host comms is no longer a direct problem for the system owner. It's all just money (which is ultimately what all this boils down to anyway). Now, their hardware CAPEX stays the same regardless if they choose A-A, or DR. Their OPEX may change (and likely will)...and this is where some of the "smoke" starts coming in related to the "Cloud". People are confusing the 'smoke' for the "Cloud" (and when I say 'smoke', I mean it in the context of "Smoke and Mirrors"!)
Note - I hate to say this, but my battery is dying and I'm going to have to stop here. But, I think you get the idea of where I am headed with my response. The Part III of this is a much larger picture, and I will get back to that, but I wanted to reply to you directly, hence this response.
I wish you could have seen my face chuckling as I was reading through all of what you wrote! And yes, I remember all of those things, in 'painful' detail...and even more. Remember how the next generation of everything was supposed to be the end-all, be-all, for everything before it, and how many times we saw those things fail. Things like ATM (asynchronous transfer mode) , and more recently UC (unified communication)? And how many evolutions have there been in database technology, or storage, or cabling. HA!...remember "CAT-7"??? That one still cracks me up! Everyone was still marveling over CAT-5"E" from the CAT-5 world, and Anixter was running around talking about their new CAT-7 product. (and before you say it...no, CAT-7 was never even a ratified standard under IEEE, it was just some trade-name Anixter came up with to sell cable, but it "sounded" cool to the uninformed!) And I forget now what the exact bandwidth and frequencies were, but I want to say they were claiming something like 750mbps and like 450MHz. I remember cracking up at some sales dude who was trying to educate this physicist (me) about it. I asked him why I would ever want 750mbps when Cisco switches jumped from 500mbps straight to 1gbps???...who would ever benefit or use it? <crickets!> LOL!!! It was like making a "revolutionary" new fire hose which was 2.5" in diameter...in a world where there were already 1" garden hoses, and 4" fire hoses (on all fire trucks) already. It made exactly zero sense! Who was ever going to use it? Not the homeowner because their spigot was only 1", and not the Fire Dept because they already pumped 10x as much water through 4" hoses. It was one of those moments that make you go..."Hmmmmm" LOL!! But I'm digressing into the weeds of techno-jargon myself (something I hope to avoid here, at least in this thread).
<Note - To the other folks reading this...this reply is not Part III of my OP, so please don't let it distract you. This reply is my reply to the esteemed EtMN, and is a bit more technical. I promise I will get to posting Part III of 'The (Dark) "Cloud" ' here shortly, so stay tuned. So please forgive my little excursion here with EtMN. You can ignore the rest of this reply unless you're into the techno stuff>
Joking aside, given your familiarity with technology, EtMN, here's something related which is germane to the OP (albeit more technical). Remember the debate between Active-Active and Disaster Recovery (DR)? This subject is one of the major factors in why people are attracted to the "cloud". In an Active-Active environment we need extreme bandwidth, far higher than any networking QOS will allow for. In fact, there can be no QOS rules at all, it has to be 100% dedicated. Of course, this means those A-A connections are going to have to be out-of-band, right? (no "networking" allowed, except for maybe dedicated hi-bandwidth Fiber-Channel (or equal) links). Well, if an entity was at that level, then they also were already Uptime Tier III (or higher) certified and practicing physical diversity. So, to get those connections they either had to install the terrestrial links themselves, or lease the dark-fiber from the Telcos. So on the one hand they had a massive CAPEX cost, or on the other they had a massive OPEX cost, AND in order to really do A-A they had to pick one or the other. THIS is why DR stayed around as long as it did (and still does). Active-Active is incomprehensibly expensive, especially over time. So, what was the next alternative? DR, right? (now hold onto that thought for a second while I hit another point). ...
So, DR is cheaper, but it's slow (like "days" compared to "seconds", kind of slow). However, DR has an interesting selling point, and we've just seen examples and read about it all over the media recently with the catastrophic Kronos debacle, right? With DR, there is nearly zero vulnerability / risk posture because (in theory anyway) your backup is off-line. So, (again, in theory) you should be able to recover from a catastrophe (unless you've contaminated your DR backup before you realized the problem). How long has Kronos been down? 4 months? How many millions of users are there (probably 250m or more)? But you see, a system like that can't be DR; it has to be A-A because of the nature of what it does (timekeeping and payroll, fundamentally). Therein lies a paradox.
Enter, the "Cloud".
Now the host --> host comms is no longer a direct problem for the system owner. It's all just money (which is ultimately what all this boils down to anyway). Now, their hardware CAPEX stays the same regardless if they choose A-A, or DR. Their OPEX may change (and likely will)...and this is where some of the "smoke" starts coming in related to the "Cloud". People are confusing the 'smoke' for the "Cloud" (and when I say 'smoke', I mean it in the context of "Smoke and Mirrors"!)
Note - I hate to say this, but my battery is dying and I'm going to have to stop here. But, I think you get the idea of where I am headed with my response. The Part III of this is a much larger picture, and I will get back to that, but I wanted to reply to you directly, hence this response.