I got that email too and I am not amused because I will not buy any additional vps from SD.
Because all I read is:
And please don't tell me you cap it just to have "same prices on all locations".
Not for that "yet another host in Chicago and Los Angeles" thing.
This has nothing to do with VPNs and if it did we'd just disable the VPN modules. I would say nearly half of our clients run VPNs while at the same time less than 20 OpenVZ clients per month use more than 100GB of bandwidth so these limits are more than enough for all but 20 clients who would need to upgrade from a $9.99/year plan to an $13.99/year plan to prevent going over their quota.
That being said, our experience has been that no matter what plans are on the nodes when we reach 50% capacity we see about 15-20Mbps of usage for that node (95%) and to give you an idea, our last bandwidth quotes in Tampa were in the $50-$200/Mbps range (with a 100Mbps commit). If a 32MB client decided to use 100% of their old bandwidth (500GB) their 1 payment per year at the old price wouldn't even cover their first month at 1Mbps (~330GB/month). These new servers have 350% more capacity so we would reach that 15-20Mbps mark without breaking a sweat. Right now if everybody on one node decided to use 100% of their bandwidth our overages would be in the $XXX range because our servers are smaller and have a lot less clients per node. If we kept the same bandwidth numbers with our new nodes we would have the potential for $XXXX in overages which would set us back a lot because of the planned expansion and server upgrades we are performing.
While I know it's selfish, after selling VPSs for 3 years our hope is for my partner and I to actually collect a paycheck by 2014 since we're currently putting in at least 40 hours a week (and with me developing a FOSS SolusVM replacement I'm putting in a lot more). Needless to say, with the amount of abuse we put up with and deal with on a daily basis the last thing we want is to have to take out a loan to pay for bandwidth overages so we've made it so even if everybody decided to do a wget to Cachefly every minute while we're sleeping we wouldn't have to worry about it.
Our contract in Tampa is up next year so we'll be able to negotiate either better bandwidth prices or find a new data center if it came down to it and when that happens we will adjust our plans accordingly.
Our intentions were not to alienate anybody. Since we sent the e-mail out, every support ticket regarding this has been filled with smiley faces that they can switch to a lower price without impacting their service. Clients love saving money without sacrificing anything and that was our goal. Nothing is set in stone, we can adjust our prices and plans as we see fit and are welcome to feedback of all kinds. If you think the bandwidth is not enough for your needs you're welcome to contact us for special pricing, but for our thousands of current clients, less than 20 of them would have been impacted by this change and the rest would save money without having to adjust their usage.
Please also take note that we are trying to work out our costs for addons (disk space, IPs, and bandwidth) so we are welcome to feedback here also. Maybe we can turn our plans into a "build you own" or "buy what you need" type of setup where the base plan comes with XGB and you can add more as needed.
Lastly, we are currently implementing a new backup solution in Denver that we will be deploying to all locations once the new hard drives arrive which we didn't discuss anywhere else but has some of an impact on bandwidth also so I figured I'd mention it here. Right now our Denver OpenVZ node is taking 6 backups per day (ever 4 hours) and is storing the last 6 4-hour backups, last 7 daily backups, last 4 weekly backups, and last 2 monthly backups which will be off-site backups once we have more backup servers to deploy.Just throwing that out there also.