CacheFly and MaxCDN, amongst others, disagree. They've built their businesses on anycast *TCP*. The idea being that routes to well-managed colocation facilities are very stable. My tests indicate both are very fast in the USA, but I suppose Gomez or some similar service would have to tell us if the percentage of dropped connections is any higher than a DNS-based CDN.
Get the devices in the hands of the reviewers who aren't paid to review and have to buy the handsets themselves. Then we'll have a better idea.
It's called Consumer Reports, right? That'd be the only source I've ever seen without bias. Every other site/mag accepts advertising, even Ars Technica.
CR's tech coverage sucks for a slashdotter, because they focus on mass-market needs. Do you have a link to a better unbiased tech reviewer?
VMware's support has been quite good for us, but then again, we paid for platinum. We actually haven't needed them too much, only for some VLAN tagging issues.
As for the licensing server bug... meh. It didn't hurt us, because we don't run on the bleeding edge version-wise. All developers screw up on occasion. We waited for the patch before upgrading to 2.5u2 plus patch at the same time, testing it in our QA cluster first. Virtualization stacks are still software that has to be managed and tested properly in your own environment.
All in all, VMware is one of the better software vendors I deal with. Leaps and bounds ahead of many.
No, I don't. We went down the VMware ESX route several years ago and have been very happy. The product has very few warts and limitations, and is rock-solid stable. Our ESX hosts stay up for months at a time.
Now, VMware's pricing leaves a lot to be desired, but when we looked into the projected operational costs of basing our virtualization project on anything else (Xen, Hyper-V, KVM), the excellent management tools and support from VMware made it the least-cost and least-risk option. We run about 70 VMs on five VMware ESX hosts.
Actually ESX is a hypervisor that runs on an old redhat distro last time I checked. Given the featureset of the newer VMWare's it would amaze me if half of them (specifically the "new" hardware support) isn't just the result of a linux kernel upgrade.
No, the VMware ESX hypervisor runs on bare metal. The management console for ESX is based on an old RedHat, but that just talks with the Hypervisor via an API. In fact, the ESXi version doesn't even have the management console, and you use the VMware client to manage the hypervisor.
But I'd appreciate a few links with recent benchmarks if you have them.
I'll see what I can dig up, but comparing VirtualBox to ESX isn't frequently done because they're so different. I recall that I saw benchamrks comparing VirtualBox to VMware desktop, and then benchmarks comparing ESX to VMware Desktop, and did the transitive analysis.
Our internal tests threw out VirtualBox (and VMware Server) as options after very simlpe IOmeter benchmarks. They were both dog-slow compared with ESX.
Yes, it is. At least it was in all of our internal testing, and every published benchmark I've ever seen.
In my experience VirtualBox beats VMWare in quite a few areas (though, sadly, not in networking). And the most recent version, 3.2, with fully asynchronous I/O widened the gap further. It's almost to the point that having virtualbox run a VM in a file on ext2 beats VMWare running the same VM with it's "partition filesystem" in normal setups.
Are you comparing VirtualBox with VMware Server or VMware Desktop? VMware ESX/vShpere are completely different products. VirtualBox may in fact out-run the lower-end, OS-hosted VMware Server and VMware Desktop. But in general, it won't come close to anything from the VMware ESX/vShpere product line (which is a hypervisor that runs on bare metal).
So, basically, if your codec does inter-frame compression, you are infringing on their patents. That doesn't sound fair to me.
Do you even understand the difference between P-frames and B-frames? Have you read any of the patents? P-frames are basically not involved, and only very specific techniques related to B-frames are in the h.264 pool. A patent on a particular B-frame coding technique does not mean you can't use any inter-frame compression.
They're not patent trolls, they're simply patent owners. The people in the MPEG-LA actually invented this shit themselves, and are quite openly following all of the rules by disclosing everything they believe applies to MPEG-4 in order to prevent patent trolling. Sure, they want to be paid for licenses, but that is very different from patent trolling.
My wife is a professional historian. Having read many of the works on her bookshelves, I can say that documentary evidence and neutrality are the absolute last priorities of a "professional" academic historian. If you actually read any of the "history" being published now, you'd know that it's all basically supposition and out-of-context pull quotes, with a focus on how a current "victim" group was mistreated so badly so long ago. "Women of the Ottoman Empire," "Being Black in Soviet Russia," "Homosexuality in Elizabethan England." They like to use the term "unwritten history", because all these groups were so oppressed they have no voice in the historical record, and so the "historians" can just make shit up that fits their worldview.
This stuff probably works 90% of the time. We lose utility power several times a year for one reason or another (usually some construction gaffe). It's a small DC in a downtown Chicago high-rise, so not a lot of the complexity of mega-facilities, but the basic pieces are the same. UPSs and/or other equipment functions as designed, and nobody notices. Who would read a blog post titled "UPS and generator not a waste of money"?
That said, so many external parties are involved and the complexity is so high that failures do happen. Our longest outage in this facility was caused by a plumber working on an another floor for another tenant. He had all the credentials to get into the right spaces, and actually fixed what he was there to fix, but also somehow managed to interrupt the redundant chilled water supply on both sides of the building at the same time.
They are being Evil. They have a perpetual obligation to keep every single feature in a time-freeze so that third parties can use them as they see fit! Ah, wait, no they don't.
Wait a minute. This is Slashdot, and the collective hive-mind knows that Microsoft does "have a perpetual obligation to keep every single feature in a time-freeze so that third parties can use them as they see fit". So does Oracle, Apple, HP, or any $big_succesful_company. Why doesn't Google?
Funny, we've been buying all-Intel servers for the same reason recently. If you factor in total system costs, the CPU cost is dominated by the aggregate of others (chassis, storage, power, network switches, cooling). It therefore makes sense to have as few servers as possible, with as much throughput per server as possible.
We actually build a simple solver for several architectures when designing our most recent server cluster, and Intel-based system had a 5-year TCO nearly 30% lower than an AMD-based solution (using 1U servers and clustered iSCSI/NAS storage) of equivalent aggregate performance. In almost all designs, top-of-rack stackable switches and distributed bottom-of-rack UPSs were TCO winners as well.
Regardless of anything else, I just have never seen any reason to keep secure, mission critical data in another companies data center. Especially email with all of its legal implications.
SaaS (or cloud or whatever buzzword you want to use) has its place. Spam filtering is a great example. Economies of scale, easy setup, reduced internal overhead. The data that flows through is not stored in any meaningful fashion.
And how, exactly, do you know that? You are completely at the mercy of the lowest-paid, most corrupt employee of your spam filtering provider. Just as you are with Gmail. They can copy and store everything if they wish. They can also turn over the keys to the G-men or lawyers after a subpoena.
If you're truly serious about protecting "secure, mission-critical data", you should be using end-to-end encryption for all messaging.
Let's see, the switches we were using had about 10us latency, of which about 1us was optical transceiver latency. That was typical a few years ago, but has since improved by an order of magnitude. I don't know where you get your "dozens of milliseconds" estimate. That certainly has not been our experience.
Chicago-to-Fairbanks=~5800 km by road, a reasonable length for a fiber path. Speed of light in fiber is ~200000 km/s. Fiber-only latency = 29 ms.
Gee, that would be too bad. But since your arguments so far are unsound, I suppose we shouldn't rush to that conclusion.
So where, exactly, are all the major datacenters in the tundra? Free colling isn't a new idea, but there still aren't many, are there?
Also, though I don't usually recommend this to people, your writing will be more credible if you turn on your spelling checker.
My last post came from my phone. Such off-topic attacks don't lend weight to your arguments.
So what you save in cooling you spend on dedicated long-haul fiber? And even if you switch to a major POP, you still have dozens of miliseconds of latency at best. You still can't support low latency operations like trading or synchronous data replication, making you customer pool thar much smaller.
That puts countries like Canada, Greenland, Denmark, Norway, Sweden and Finland in high demand for datacenters (end technicians to staff them). If the US didn't have rediculous data laws, Alska might also be ideal.
Some datacenters perhaps, that don't need good Internet connectivity. But the latency between major populations and the far North makes those locations less desirable. We have struggles with the latency between Chicago and Dallas with some applications; Chicago-to-Fairbanks would be quite a bit more painful.
Your facts are wrong. Microsoft makes lots and lots of money from other things. Is server products are a cash cow by any measure, most notably SQL Server, SharePoint, and Exchange. In fact, these products made about $1.25 billion in the last 3 months. This "small" piece of Microsoft's business is far larger than 99.9% of the technology companies out there, and is even 50% of the size of the entirety of Google.
Add up the value of the man-hours spent on the issue so far, plus the costs associated with your downtime and performance issues, and the transition may look a lot more attractive. You probably cost your company at least US$100/hr including overhead and benefits.
Slashdot Mirror
CacheFly and MaxCDN, amongst others, disagree. They've built their businesses on anycast *TCP*. The idea being that routes to well-managed colocation facilities are very stable. My tests indicate both are very fast in the USA, but I suppose Gomez or some similar service would have to tell us if the percentage of dropped connections is any higher than a DNS-based CDN.
Get the devices in the hands of the reviewers who aren't paid to review and have to buy the handsets themselves. Then we'll have a better idea.
It's called Consumer Reports, right? That'd be the only source I've ever seen without bias. Every other site/mag accepts advertising, even Ars Technica.
CR's tech coverage sucks for a slashdotter, because they focus on mass-market needs. Do you have a link to a better unbiased tech reviewer?
They could provide the source code by any means they wished. Why would they have to provide it "through the App store"? Read the license.
VMware's support has been quite good for us, but then again, we paid for platinum. We actually haven't needed them too much, only for some VLAN tagging issues.
As for the licensing server bug... meh. It didn't hurt us, because we don't run on the bleeding edge version-wise. All developers screw up on occasion. We waited for the patch before upgrading to 2.5u2 plus patch at the same time, testing it in our QA cluster first. Virtualization stacks are still software that has to be managed and tested properly in your own environment.
All in all, VMware is one of the better software vendors I deal with. Leaps and bounds ahead of many.
I think it was Vmware ESX 3.5 (u1 I think) and VirtualBox 2.0 on Windows X64. It was mid-2008, so it was whatever was current then.
Do you work for VMWARE?
No, I don't. We went down the VMware ESX route several years ago and have been very happy. The product has very few warts and limitations, and is rock-solid stable. Our ESX hosts stay up for months at a time.
Now, VMware's pricing leaves a lot to be desired, but when we looked into the projected operational costs of basing our virtualization project on anything else (Xen, Hyper-V, KVM), the excellent management tools and support from VMware made it the least-cost and least-risk option. We run about 70 VMs on five VMware ESX hosts.
Actually ESX is a hypervisor that runs on an old redhat distro last time I checked. Given the featureset of the newer VMWare's it would amaze me if half of them (specifically the "new" hardware support) isn't just the result of a linux kernel upgrade.
No, the VMware ESX hypervisor runs on bare metal. The management console for ESX is based on an old RedHat, but that just talks with the Hypervisor via an API. In fact, the ESXi version doesn't even have the management console, and you use the VMware client to manage the hypervisor.
But I'd appreciate a few links with recent benchmarks if you have them.
I'll see what I can dig up, but comparing VirtualBox to ESX isn't frequently done because they're so different. I recall that I saw benchamrks comparing VirtualBox to VMware desktop, and then benchmarks comparing ESX to VMware Desktop, and did the transitive analysis.
Our internal tests threw out VirtualBox (and VMware Server) as options after very simlpe IOmeter benchmarks. They were both dog-slow compared with ESX.
Is ESX so much faster than VirtualBox?
Yes, it is. At least it was in all of our internal testing, and every published benchmark I've ever seen.
In my experience VirtualBox beats VMWare in quite a few areas (though, sadly, not in networking). And the most recent version, 3.2, with fully asynchronous I/O widened the gap further. It's almost to the point that having virtualbox run a VM in a file on ext2 beats VMWare running the same VM with it's "partition filesystem" in normal setups.
Are you comparing VirtualBox with VMware Server or VMware Desktop? VMware ESX/vShpere are completely different products. VirtualBox may in fact out-run the lower-end, OS-hosted VMware Server and VMware Desktop. But in general, it won't come close to anything from the VMware ESX/vShpere product line (which is a hypervisor that runs on bare metal).
So, basically, if your codec does inter-frame compression, you are infringing on their patents. That doesn't sound fair to me.
Do you even understand the difference between P-frames and B-frames? Have you read any of the patents? P-frames are basically not involved, and only very specific techniques related to B-frames are in the h.264 pool. A patent on a particular B-frame coding technique does not mean you can't use any inter-frame compression.
They're not patent trolls, they're simply patent owners. The people in the MPEG-LA actually invented this shit themselves, and are quite openly following all of the rules by disclosing everything they believe applies to MPEG-4 in order to prevent patent trolling. Sure, they want to be paid for licenses, but that is very different from patent trolling.
...when I completely exit chrome and re-visit the same site.
So... you agree word-for-word with what I wrote, but I'm a complete and utter twat?
Petty power - like that wielded by network administrators... corrupts immensely out of all proportion to the actual power.
What's good for the goose is good for the gander.
The question is, a beacon indicating exactly what?
"Get the hell out of our way! We're not real sure where the brakes are on this thing, and we've been drinking."
My wife is a professional historian. Having read many of the works on her bookshelves, I can say that documentary evidence and neutrality are the absolute last priorities of a "professional" academic historian. If you actually read any of the "history" being published now, you'd know that it's all basically supposition and out-of-context pull quotes, with a focus on how a current "victim" group was mistreated so badly so long ago. "Women of the Ottoman Empire," "Being Black in Soviet Russia," "Homosexuality in Elizabethan England." They like to use the term "unwritten history", because all these groups were so oppressed they have no voice in the historical record, and so the "historians" can just make shit up that fits their worldview.
This stuff probably works 90% of the time. We lose utility power several times a year for one reason or another (usually some construction gaffe). It's a small DC in a downtown Chicago high-rise, so not a lot of the complexity of mega-facilities, but the basic pieces are the same. UPSs and/or other equipment functions as designed, and nobody notices. Who would read a blog post titled "UPS and generator not a waste of money"?
That said, so many external parties are involved and the complexity is so high that failures do happen. Our longest outage in this facility was caused by a plumber working on an another floor for another tenant. He had all the credentials to get into the right spaces, and actually fixed what he was there to fix, but also somehow managed to interrupt the redundant chilled water supply on both sides of the building at the same time.
They are being Evil. They have a perpetual obligation to keep every single feature in a time-freeze so that third parties can use them as they see fit! Ah, wait, no they don't.
Wait a minute. This is Slashdot, and the collective hive-mind knows that Microsoft does "have a perpetual obligation to keep every single feature in a time-freeze so that third parties can use them as they see fit". So does Oracle, Apple, HP, or any $big_succesful_company. Why doesn't Google?
Funny, we've been buying all-Intel servers for the same reason recently. If you factor in total system costs, the CPU cost is dominated by the aggregate of others (chassis, storage, power, network switches, cooling). It therefore makes sense to have as few servers as possible, with as much throughput per server as possible.
We actually build a simple solver for several architectures when designing our most recent server cluster, and Intel-based system had a 5-year TCO nearly 30% lower than an AMD-based solution (using 1U servers and clustered iSCSI/NAS storage) of equivalent aggregate performance. In almost all designs, top-of-rack stackable switches and distributed bottom-of-rack UPSs were TCO winners as well.
Regardless of anything else, I just have never seen any reason to keep secure, mission critical data in another companies data center. Especially email with all of its legal implications. SaaS (or cloud or whatever buzzword you want to use) has its place. Spam filtering is a great example. Economies of scale, easy setup, reduced internal overhead. The data that flows through is not stored in any meaningful fashion.
And how, exactly, do you know that? You are completely at the mercy of the lowest-paid, most corrupt employee of your spam filtering provider. Just as you are with Gmail. They can copy and store everything if they wish. They can also turn over the keys to the G-men or lawyers after a subpoena.
If you're truly serious about protecting "secure, mission-critical data", you should be using end-to-end encryption for all messaging.
Let's see, the switches we were using had about 10us latency, of which about 1us was optical transceiver latency. That was typical a few years ago, but has since improved by an order of magnitude. I don't know where you get your "dozens of milliseconds" estimate. That certainly has not been our experience.
Chicago-to-Fairbanks=~5800 km by road, a reasonable length for a fiber path. Speed of light in fiber is ~200000 km/s. Fiber-only latency = 29 ms.
Gee, that would be too bad. But since your arguments so far are unsound, I suppose we shouldn't rush to that conclusion.
So where, exactly, are all the major datacenters in the tundra? Free colling isn't a new idea, but there still aren't many, are there?
Also, though I don't usually recommend this to people, your writing will be more credible if you turn on your spelling checker.
My last post came from my phone. Such off-topic attacks don't lend weight to your arguments.
So what you save in cooling you spend on dedicated long-haul fiber? And even if you switch to a major POP, you still have dozens of miliseconds of latency at best. You still can't support low latency operations like trading or synchronous data replication, making you customer pool thar much smaller.
That puts countries like Canada, Greenland, Denmark, Norway, Sweden and Finland in high demand for datacenters (end technicians to staff them). If the US didn't have rediculous data laws, Alska might also be ideal.
Some datacenters perhaps, that don't need good Internet connectivity. But the latency between major populations and the far North makes those locations less desirable. We have struggles with the latency between Chicago and Dallas with some applications; Chicago-to-Fairbanks would be quite a bit more painful.
IE and Windows media Player have respictively done the same thingsince the Windows NT 4 era. So?
Your facts are wrong. Microsoft makes lots and lots of money from other things. Is server products are a cash cow by any measure, most notably SQL Server, SharePoint, and Exchange. In fact, these products made about $1.25 billion in the last 3 months. This "small" piece of Microsoft's business is far larger than 99.9% of the technology companies out there, and is even 50% of the size of the entirety of Google.
Add up the value of the man-hours spent on the issue so far, plus the costs associated with your downtime and performance issues, and the transition may look a lot more attractive. You probably cost your company at least US$100/hr including overhead and benefits.