True bandwidth (throughput)

[quarghost]

We're all aware of our monthly bandwidth limits. But does anybody know what the max throughput to the server is?

I assume all the servers are residing on gigabit switches, but that does not mean our data hits the internet at gigabit speeds. I'm wondering what this is capped at per server.

For example, a rather large Pittsburgh based collocation provider with a name that sounds like 'chair' offers three packages. Their base packages offer between 2 and 5Mbps to the rack. Now obviously we're not collocating here so things are a little different, but I was wondering what PVPS does in this regard.

I don't have a speed issue, I was just wondering for curiosity sake.

Also, I know this question is better addressed by going directly to pre-sales, but I figured I'd open it up to the masses so that we might hear comments about customers perceived speed and experience.

[Daniel]

Originally Posted by quarghost

We're all aware of our monthly bandwidth limits. But does anybody know what the max throughput to the server is?

I assume all the servers are residing on gigabit switches, but that does not mean our data hits the internet at gigabit speeds. I'm wondering what this is capped at per server.

For example, a rather large Pittsburgh based collocation provider with a name that sounds like 'chair' offers three packages. Their base packages offer between 2 and 5Mbps to the rack. Now obviously we're not collocating here so things are a little different, but I was wondering what PVPS does in this regard.

I don't have a speed issue, I was just wondering for curiosity sake.

Also, I know this question is better addressed by going directly to pre-sales, but I figured I'd open it up to the masses so that we might hear comments about customers perceived speed and experience.

Hi Michael -

Each hardware node is connected to a 100meg switch port.

This helps us to ensure that if, during an attack or the like, the entire aggregation switch does not get completely saturated and it also helps to limit the amount of bandwidth overage you could be hit with.

99% of our VPS customers will never need even close to 100 meg.

Hope that helps answer your question.

[quarghost]

Thanks for answering.

Originally Posted by Daniel

Each hardware node is connected to a 100meg switch port.

Doesn't this just apply to the internal network? I'm trying to find out about the thoroughput.

Here's a poor analogy: I have a DSL connection at home that I run a few servers off of. I get 20Mbps down and roughly 1.3Mbps up (the 20 figure due to my proximity to the CO, but that's neither here nor there). My internal network is 100Mbps. So I understand that connections between servers (internally) and to the DSL modem will max out near 100Mbps. But obviously this doesn't mean that traffic coming in will do so any higher than 20Mbps and traffic going out than 1.3Mbps.

Now all that is pipe speed. Real speed is obviously going to be lower. And this doesn't account for traffic shaping (i.e. I could set my switch to only allow 10Mbps of traffic between ports, which would then theoretically lower the perceived bandwidth to the server.

So my question relating to PowerVPS is: What up/down rate are we hitting the internet with? We're 100Mbps to the switch, and that's pretty standard, but what's the limited rate on the other side of the switch? I say limited because I assume you're capping this as well for the same reasons you would internally.

Not that this makes any real difference one way or the other, as much of this stuff has more to do with perception of speed than reality, but when people ask me what speed the connection is, I'd like to say:______.

[Daniel]

Originally Posted by quarghost

Thanks for answering.



Doesn't this just apply to the internal network? I'm trying to find out about the thoroughput.

Here's a poor analogy: I have a DSL connection at home that I run a few servers off of. I get 20Mbps down and roughly 1.3Mbps up (the 20 figure due to my proximity to the CO, but that's neither here nor there). My internal network is 100Mbps. So I understand that connections between servers (internally) and to the DSL modem will max out near 100Mbps. But obviously this doesn't mean that traffic coming in will do so any higher than 20Mbps and traffic going out than 1.3Mbps.

Now all that is pipe speed. Real speed is obviously going to be lower. And this doesn't account for traffic shaping (i.e. I could set my switch to only allow 10Mbps of traffic between ports, which would then theoretically lower the perceived bandwidth to the server.

So my question relating to PowerVPS is: What up/down rate are we hitting the internet with? We're 100Mbps to the switch, and that's pretty standard, but what's the limited rate on the other side of the switch? I say limited because I assume you're capping this as well for the same reasons you would internally.

Not that this makes any real difference one way or the other, as much of this stuff has more to do with perception of speed than reality, but when people ask me what speed the connection is, I'd like to say:______.

Hi Michael -

I hope I understand you correctly, but if this does not answer your question please let me know.

Each physical hardware node (that you share with X number of other customers) is connected to our public (not private internal traffic) switches on a 100meg port.

Theoretically, if no one else on the hardware node where your VPS resides is using any bandwidth at all, your single VPS could push 100meg of traffic out to the Internet or in depending on what is happening.

We do not cap each individual VPS, only the hardware node. Keep in mind that pushing 100meg consistently will cause a good amount of bandwidth to be pushed and likely resulting in a bandwidth overage bill.

Customers have used the full 100meg port before and (on the PowerVPS side of things anyway) this is usually due to a compromised VPS sending out DDoS attacks, or a VPS that is being attacked.

I hope that helps out.