This is what I do know: 1. Well Facebook does use servers with a height of 1.5 rack units. So they can use larger fans (which obviously need less rotations)
3. non of the providers which did their own custom servers choose DC. AFAIK only Facebook uses some DC. The servers have a AC and DC input. The DC input is connected to the rack with UPS which keeps the other 2 racks (in the set of 3 racks) powered in case the AC is out.
This policy does not work because one of the first website people visit is Google, they change their certificated on a weekly basis (yes all servers; every week they roll out new certs).
It rather use PowerDNS, as it supports DNSSEC. It also uses PostgreSQL a superior solution in many situations. PowerDNS supports many backends it can also serve from bind zone files for example.
Have a good look at what Google and Facebook are doing and how Facebook is very open about it and collaborating with others in the OpenCompute project.
Companies like HP and Dell are looking very closely at what they can use from these designs to build servers for the rest of us. I think Dell is even one of the members of the Open Compute project.
The most important "innovation" if you ask me is to close off the hot corridore and have all the connectors at the front of the server in the cold corridore and make it possible to do all maintaince from front too. Then you use free cooling for the cold corridore with higher temperatures then most people use now.
The hot corridore will get really hot, but you will never have to enter it (other than maybe replace a ceiling fan or something like that).
In the Wired article I think Google mentions, no1 enters the hot corridor unless all servers in the rows for that corridore are turned off. That makes it pretty clear how hot it is and it also shows you where that hot air is going directly out of the datacenter in the open air.
It is just that the modern world looks different than it did all those years ago.
The choice for the number of bits an in IPv4 address wasn't wrong at the time, but we need mode bits if we want to Internet to keep growing.
It's the same with the GPUs of today, the best way to talk to them is not the way X11 works.
Compositing seems to be what the creators of Wayland want to do.
Sure you can do compositing with X11, but it isn't ideal. But compositing is what most systems already do. Compiz and KWin, GNOME3 and Unity all are doing compositing I believe.
I think i read newer versions of Windows (since Vista ?) and Mac OS X also do compositing.
If compositing is all you ever do, you don't need everything else from X11. Especially if that abstraction/code is in the "hotpath", the code paths which are the most performance critical.
I forgot to mention code review. When it's open source other people can look at the code, if they do is a totally different question. But it is possible. With bigger projects, I believe they do.
I'm just glad when a software vendor releases a fix, including security, it only takes up to a couple of days until my system gets it updates.
Everything else just means: you need to have fait in the original programmer and the team that handles the vulnerability reports.
Open source or not.
I believe open source works better though, I've never seen that someone reported a security bug was delayed for months on end.
Other then something like this: "Last year (2011) there was a period of several months when the CentOS project did not issue any security advisories or updates for CentOS 6. Many CentOS users got frustrated and worried about their system security"
Which just means people have the choice to replace CentOS with an other distribution and mostly life happily ever after.
With a closed system you can't, there is only one vendor of Windows, right ?
Actually, even though browsers get security fixes every release.
They are usually one of the parts of the system that do get these updates regularly.
They deal with input from the open Internet on a daily basis, they have to be updated.
It has been very clear the last few years that most drive-by-infections happen through plugins.
But the plugins don't get the updates. So what are browser developers supposed to do ? Firefox developers would like to prevent drive-by-infections, because they are the most common way virusses get on to computers now a days (for a couple of years now).
So they (temporarily) disable the plugins that have known security problems.
Slashdot Mirror
This is what I do know:
1. Well Facebook does use servers with a height of 1.5 rack units. So they can use larger fans (which obviously need less rotations)
2. OVH which says they own and operate the largest datacenter in the world do use water cooling: https://www.youtube.com/watch?v=4e97g7_qSxA http://www.ovh.co.uk/dedicated_servers/hg_2012_watercooling.xml
3. non of the providers which did their own custom servers choose DC. AFAIK only Facebook uses some DC. The servers have a AC and DC input. The DC input is connected to the rack with UPS which keeps the other 2 racks (in the set of 3 racks) powered in case the AC is out.
I don't know that is what I heared in a presentation, I think it is somewhere online. I'll see if I can find it somewhere.
I'm sure that would work really well: NOT
This policy does not work because one of the first website people visit is Google, they change their certificated on a weekly basis (yes all servers; every week they roll out new certs).
It rather use PowerDNS, as it supports DNSSEC. It also uses PostgreSQL a superior solution in many situations. PowerDNS supports many backends it can also serve from bind zone files for example.
Have a good look at what Google and Facebook are doing and how Facebook is very open about it and collaborating with others in the OpenCompute project.
Companies like HP and Dell are looking very closely at what they can use from these designs to build servers for the rest of us. I think Dell is even one of the members of the Open Compute project.
The most important "innovation" if you ask me is to close off the hot corridore and have all the connectors at the front of the server in the cold corridore and make it possible to do all maintaince from front too. Then you use free cooling for the cold corridore with higher temperatures then most people use now.
The hot corridore will get really hot, but you will never have to enter it (other than maybe replace a ceiling fan or something like that).
In the Wired article I think Google mentions, no1 enters the hot corridor unless all servers in the rows for that corridore are turned off. That makes it pretty clear how hot it is and it also shows you where that hot air is going directly out of the datacenter in the open air.
Some say ARM servers are the future:
http://www.youtube.com/watch?v=njmQBqUuYqU
Yes, they archive on tape.
I wouldn't be surprised if Amazon does it too:
http://aws.amazon.com/glacier/
Seems they measure differently though:
http://gigaom.com/cloud/whose-data-centers-are-more-efficient-facebooks-or-googles/
What I found interresting is that the PUE at that facility is 1.10, the Facebook Pineville datacenter has a PUE of 1.07
I believe that in many fields it's true that the more you know about a subject, the more you realize how little that really is.
firewalls and routers ? Really ? Only if they use ASICs.
It's not that the legacy is crap.
It is just that the modern world looks different than it did all those years ago.
The choice for the number of bits an in IPv4 address wasn't wrong at the time, but we need mode bits if we want to Internet to keep growing.
It's the same with the GPUs of today, the best way to talk to them is not the way X11 works.
Compositing seems to be what the creators of Wayland want to do.
Sure you can do compositing with X11, but it isn't ideal. But compositing is what most systems already do. Compiz and KWin, GNOME3 and Unity all are doing compositing I believe.
I think i read newer versions of Windows (since Vista ?) and Mac OS X also do compositing.
If compositing is all you ever do, you don't need everything else from X11. Especially if that abstraction/code is in the "hotpath", the code paths which are the most performance critical.
Maybe that is why they are in the S&P 500, they sell to management in a language they understand ?
They probably felt like they needed to do a rework anyway.
That they are going to be right in the end too.
Open systems, protocols and standards should be the norm. Preferable open source as well.
Isn't that why a good RAID system has a hot spare drive installed (the spare hot spare did not get all the writes the others did).
A lot of people outside Google believe that ChromeOS was actually their ultimate goal, but they needed to release a smartphone OS before it was ready.
A HTML5 video-tag would have been fine too.
Hey, I'm just saying that is the only choice you have. Other then, use a different software/other developers. Or build it yourself of course.
They are fast on updates, it just is security bugs aren't found early enough.
Actually, you should.
An organisation like the FBI just takes a bunch of servers any time they like:
http://bits.blogs.nytimes.com/2011/06/21/f-b-i-seizes-web-servers-knocking-sites-offline/
If they are from the right tenant does not really matter to them. It will take many months before your server is returned.
I forgot to mention code review. When it's open source other people can look at the code, if they do is a totally different question. But it is possible. With bigger projects, I believe they do.
I'm just glad when a software vendor releases a fix, including security, it only takes up to a couple of days until my system gets it updates.
Everything else just means: you need to have fait in the original programmer and the team that handles the vulnerability reports.
Open source or not.
I believe open source works better though, I've never seen that someone reported a security bug was delayed for months on end.
Other then something like this: "Last year (2011) there was a period of several months when the CentOS project did not issue any security advisories or updates for CentOS 6. Many CentOS users got frustrated and worried about their system security"
Which just means people have the choice to replace CentOS with an other distribution and mostly life happily ever after.
With a closed system you can't, there is only one vendor of Windows, right ?
I'm fine with what Google has done and it is also the right of these people to protest against them doing so.
There is nothing wrong here, everything is fine, all is great, maybe not everyone is happy (about everything), but that's just life.
Actually, even though browsers get security fixes every release.
They are usually one of the parts of the system that do get these updates regularly.
They deal with input from the open Internet on a daily basis, they have to be updated.
It has been very clear the last few years that most drive-by-infections happen through plugins.
But the plugins don't get the updates. So what are browser developers supposed to do ? Firefox developers would like to prevent drive-by-infections, because they are the most common way virusses get on to computers now a days (for a couple of years now).
So they (temporarily) disable the plugins that have known security problems.
Is that so strange ?
Official 64-bit builds for Linux have been available for atleast a year if not years.
Nightly 64-bit build of for Windows have been available for atleast a half a year if not a full year.