What are the key (practical) differences between Cloud VPS and ‘normal’ VPS?
]
Answer #2
Yes, that makes sense Ben. Thank you.
How about dynamic workload or general capacity elasticity? Is there anything about the cloud solution that allows for real-time handling of abnormal peaks – say for example I were running a contest/giveaway and during the hour of publishing the results I have 1,000 people logged into my phpBB forum instead of the typical 100-120. A normal hosting solution would fail on memory or exceed the allowable concurrent mySQL connections.
Is there anything special about Cloud that would handle a blip like that? Or is it strictly enhanced hardware redundancy/fail-over?
Answer #3
No problem!
To gain the general capacity elasticity you really need to build an application that is designed to use the cloud. Where as this is a virtualized server living in the cloud and in all other regards a normal server (I’ll answer the raw scaling q further down).
The marketing buzz around cloud really makes it seem magic, really it just is an extra layer of hardware protection for most customers. It really only get’s magical if you build an application for the specific cloud layer so that technical limits on server capacity disappear. Or if you need a normal server than changes it’s resources daily/hourly.
So to explain specifically:
It is slightly easier to upgrade in Cloud but since we handle that entire process in either platform it’s not something you can easily see the benefit of. Down the road as we roll out unmanaged vps in both platforms the ability to scale and move around will become more available to clients of course. In most cases that still takes a reboot to upgrade/downgrade a server to a new level just due to limits on that virtualized environment.
So in terms of your specific example, you could upgrade before that rush comes and then downgrade after it’s over. Currently we view that on a monthly basis though and not daily / hourly (we are looking at offering daily/hourly billing though). So in this specific case it would not help you handle that sudden jump any better.
One more thing that is coming down the road is application specific hosting built on the cloud. In that case something like this is possible where you are buying a phpbb application from us and get only access to it and no server level. Then we would design it to really take advantage of the cloud and a cluster of servers. We have no plans currently to develop that for phpbb, but we do for WordPress.
Let me know if you have any questions,
Thanks, Ben
]Answer this Question
You must be Logged In to post an Answer.
Not a member yet? Sign Up Now »
Answer #1
Sure the only different is having your VPS in the Cloud gives it an extra layer of hardware redundancy.
In the cloud if your server has a motherboard failure, ram failure, or similar it would move to a different server in 1 to 10 minutes.
On a normal vps if there was a hardware failure we would need to bring the server down, replace the hardware and get it back up. Plus if the raid array for storage was lost we would need to do a full restoer which would take time (very rare though).
Does that make sense?
Hardware failure is rare already, it just depends if you want to make it ever rarer.
Thanks, Ben