3 ms·

It seems like transfer would be an 'easy' upgrade in that it doesn't require a lot of per-server manual effort, just a change to the billing system (threshold a

It seems like transfer would be an 'easy' upgrade in that it doesn't require a lot of per-server manual effort, just a change to the billing system (threshold at which accounts are billed for overages). But I dunno if that makes sense, given purchasing decisions and your cost structure. Are people's purchase decisions keying off transfer quota at all, or are RAM/HDD are the real price-point issues? And if they are, would doubling transfer, say, make a difference, or would you have to match Linode (currently offering 10x the transfer at equivalent price points)? So maybe this is a dumb idea.

lsc 13y ago
oh, you are right on all above. right now I'm way overbuying transit... not quite 10x, but 5x, so I really should increase quotas. For a long time, it was that my network was so shitty that doubling traffic would make the thing fall over, but that problem is fixed... new network is way more robust. so yeah, I should increase transfer expectations. I mean, in reality, I follow the 'use all you want, limit you when you get crazy' model, rather than hitting you with a big bill when you go over. (I should make that a selectable option) so it's really similar to what digital ocean is doing with their 'unlimited' - obviously, if you go crazy on an 'unlimited' system, you get limited. My limits are 'setting expectations' - but right now they are setting expectations that would have been reasonable in 2007, which is silly, as my current network is pretty robust, and really can handle more.