HyperV Nested Virtualization Announced for Container Support
-
@DustinB3403 said:
I don't use Azure (XenServer is my preference), the one Hyper-V server was completely setup ass backwards by our MSP, and is in production, so yeah, lets not go down that road.
Portable, how so? Because you can simply export the primary VM, which includes all other VM's configured on it?
Or am I missing your point?
Yes, instead of moving VM by VM from one place to another, you can move your datacenter from one place to another.
-
@DustinB3403 said:
Well I guess that makes sense. But if you have an Azure solution, why would you also host the solution on site?
In the example, Azure is the DR site. Lots of companies want the speed (network latency) and cost of an on premises solution but cannot afford a DR facility. Azure is that DR facility. This is actually a major talking point of Azure sales. They work hard to make this a viable option for their offerings.
-
@DustinB3403 said:
Why not just leave it on Azure. I'm going out on a limb and thinking Microsoft has a much more robust and reliable data center then most businesses.
Indeed they do, our recent outages aside, but it isn't cheap and if your issues are primarily network based or you have need for fast networks or you just don't need that robustness you can offset it this way.
-
Is the cost of being able to use Azures service something that can be "enabled" with an email? I would assume it to be a monthly reoccurring cost.
And does the price get reduced if you aren't actively using it, but have the option to use it on a whim?
-
@DustinB3403 said:
Is the cost of being able to use Azures service something that can be "enabled" with an email? I would assume it to be a monthly reoccurring cost.
Like all IaaS cloud, you pay by them minute or hour for capacity. So when not used, you don't pay.
-
@DustinB3403 said:
And does the price get reduced if you aren't actively using it, but have the option to use it on a whim?
That's the definition of cloud
-
Azure, Amazon AWS, Rackspace, Softlayer, Digital Ocean, Vultr, etc. are all the same. Capacity is by minute or hour (you can choose larger blocks with AWS for discounts) and so you literally pay only for what you use. The idea of cloud is that you can elastically grow or contract as your load changes so that you are paying for only what you need rather than for your maximum needs all of the time.
-
Ah well at least it's cost-viable, rather than being a flat rate minimum per month plus usage.
Sorry it's still early, not thinking very clearly.
-
@DustinB3403 said:
Ah well at least it's cost-viable, rather than being a flat rate minimum per month plus usage.
Sorry it's still early, not thinking very clearly.
Very cost-viable. The issues are that you need to get that infrastructure up there in the case of an outage and that could be a bit of a problem since you presumably are offline and your entire infrastructure might be pretty big.
-
So one has to ask, if you have a site wide unplanned outage but do have access to Azure, how do you get your Hyper-V data to them so you can at least get to a functional state?
Just play out the different "DR scenarios" and options. As I'm already thinking of 6 different questions.
- How do you get your hyper-v infrastructure to Azure in the event of an unplanned outage?
- Do they have some solution to keep a "ready-to-go" state of your hyper-v network/
-
@DustinB3403 said:
So one has to ask, if you have a site wide unplanned outage but do have access to Azure, how do you get your Hyper-V data to them so you can at least get to a functional state?
Generally you would keep it synced there all of the time.
-
@DustinB3403 said:
- Do they have some solution to keep a "ready-to-go" state of your hyper-v network/
You can copy to cold storage there.
-
So question 3 how do "you" work if you have an unplanned outage, even if Azure is running your VM Fleet?
This is obviously a rhetorical question, and doesn't need an answer.
-
That sounds like a great solution then, I wonder what the pricing options are, gonna do some research.
-
@DustinB3403 said:
So question 3 how do "you" work if you have an unplanned outage, even if Azure is running your VM Fleet?
This is obviously a rhetorical question, and doesn't need an answer.
I'll offer an answer anyway.
Grab your laptop and go home... or to the nearest place with internet access. 
-
@DustinB3403 said:
That sounds like a great solution then, I wonder what the pricing options are, gonna do some research.
Not sure if this is what you are looking for:
-
https://azure.microsoft.com/en-us/pricing/details/storage/ specifically.
But thanks
-
Here is the pricing for site recovery. https://azure.microsoft.com/en-us/pricing/details/site-recovery/.
Is an instance an individual virtual machine? If that is the case the pricing isn't really that bad. Looks like you will also incur charges for storage transactions and usage.
-
You pay for all usage, but only for usage not potential.
-
@dafyre said:
@DustinB3403 I would think more portable as in live migration... You don't have any generators at your main site and the power company needs to take you down for 24 hours. You could then Migrate your Nested Hyper-V infrastructure to Azure for a day and not have to scramble to find generators and such, and you don't even have to scramble to bring things back up because they never go down.
That would be the way I see it. As I said, I may be completely out of my mind, but that is what I envision with containers.
We'll ignore the completely badly configured Hyper-V setup that you have now.
How small are your VMs that you could push them to Azure in less than 24 hours?
