EMC is one of the leading manufacturers of high-end data storage systems for mainframe computers and I worked with their equipment and their people for years. They're exceptionally talented, great people, and they provide the kit you need for building big systems.
Dell has just bought them (The Guardian: Dell to buy EMC in $67bn deal)
The first thing you know off the top is Dell is pulling out of making personal computers as they have focused everything on Windows ... but Windows isn't focused on much of anything anymore. Therefore they make the move to the big league.
Maybe more interesting is who did not make a move. EMC is a plum of an acquisition yet IBM did not buy them and they would benefit most of all. HP did not make a move because they're gutted by what Carly Fiorina did to them. Cisco is another big league player but they don't have the capital to make such a move.
It's an unorthodox but brilliant move by Dell as they had no more prospect than waiting for death previously but this opens the world back up to them.
Some friends worked for EMC and probably still do. I know many IT types have invested in EMC stock. This likely is good news all over the place.
This was some of the coolest kit I ever saw in mainframes. The big mainframes aren't that cool because, first, they aren't big. A mainframe vastly more powerful than the giant machines on which I worked in the early days is now about the size of a refrigerator. They also don't do anything that's worth watching. They do fantastic things inside their clouds but no-one can see that except by visualizing it in whatever twisted way works for you.
All you'll ever see of computers in sci-fi movies is blinking LEDs and spinning tape drives ... because they don't do anything else you can watch.
What was cool about EMC was designing. We've got our primary disks here, hundreds of them, and these are copied six or eight times so there is always data recovery. In a personal system that would be enough but it's not good enough for big ones.
There's your primary storage but that needs even more backup because what do you do if the entire primary site is wiped out. Not only is your infrastructure smashed all to hell but all of your records are gone as well. That is one deceased parrot.
So you need another site for yet more backups. This one has to be close because you need fast data transmission but it mustn't be so close it can be affected by whatever took out the primary site. For example, if your primary site is in Los Angeles, putting your secondary site in San Francisco won't do you much good because whatever took out the primary in L.A. will probably also take out the secondary in San Francisco.
But that's not enough either. Maybe it's a situation with a Hurricane Sandy and the secondary site really does get taken out at the same time as the primary.
There's nothing else for it. There needs to be a 'bunker' site in a location which is geographically-stable (i.e. little risk of earthquakes) and is generally isolated. Maybe you think of somewhere like Nebraska and that's fine. This is the 'long haul' target site and this is the archive.
All of that hardware is interesting and designing the targets for these things is interesting as well but the cool part is making it work. That gets lots of little boxes and circles with arrows between them on a drawing board and we all go ooh ah at the sophistication of something we design all the while knowing not one person will ever see it.
That's ok. We did. And that was just fine.
Dell has just bought them (The Guardian: Dell to buy EMC in $67bn deal)
The first thing you know off the top is Dell is pulling out of making personal computers as they have focused everything on Windows ... but Windows isn't focused on much of anything anymore. Therefore they make the move to the big league.
Maybe more interesting is who did not make a move. EMC is a plum of an acquisition yet IBM did not buy them and they would benefit most of all. HP did not make a move because they're gutted by what Carly Fiorina did to them. Cisco is another big league player but they don't have the capital to make such a move.
It's an unorthodox but brilliant move by Dell as they had no more prospect than waiting for death previously but this opens the world back up to them.
Some friends worked for EMC and probably still do. I know many IT types have invested in EMC stock. This likely is good news all over the place.
This was some of the coolest kit I ever saw in mainframes. The big mainframes aren't that cool because, first, they aren't big. A mainframe vastly more powerful than the giant machines on which I worked in the early days is now about the size of a refrigerator. They also don't do anything that's worth watching. They do fantastic things inside their clouds but no-one can see that except by visualizing it in whatever twisted way works for you.
All you'll ever see of computers in sci-fi movies is blinking LEDs and spinning tape drives ... because they don't do anything else you can watch.
What was cool about EMC was designing. We've got our primary disks here, hundreds of them, and these are copied six or eight times so there is always data recovery. In a personal system that would be enough but it's not good enough for big ones.
There's your primary storage but that needs even more backup because what do you do if the entire primary site is wiped out. Not only is your infrastructure smashed all to hell but all of your records are gone as well. That is one deceased parrot.
So you need another site for yet more backups. This one has to be close because you need fast data transmission but it mustn't be so close it can be affected by whatever took out the primary site. For example, if your primary site is in Los Angeles, putting your secondary site in San Francisco won't do you much good because whatever took out the primary in L.A. will probably also take out the secondary in San Francisco.
But that's not enough either. Maybe it's a situation with a Hurricane Sandy and the secondary site really does get taken out at the same time as the primary.
There's nothing else for it. There needs to be a 'bunker' site in a location which is geographically-stable (i.e. little risk of earthquakes) and is generally isolated. Maybe you think of somewhere like Nebraska and that's fine. This is the 'long haul' target site and this is the archive.
All of that hardware is interesting and designing the targets for these things is interesting as well but the cool part is making it work. That gets lots of little boxes and circles with arrows between them on a drawing board and we all go ooh ah at the sophistication of something we design all the while knowing not one person will ever see it.
That's ok. We did. And that was just fine.
No comments:
Post a Comment