Setting up a Tor relay on a VPS/cloud is a relatively straightforward process, which is great for people who enjoy that warm-fuzzy-i'm-making-a-difference feeling we get by contributing bandwidth to the network. However, most providers have a monthly limit on how much bandwidth you can contribute before charging extra, and that's something most of us would like to avoid.
torrc includes a really handy setting called
AccountingMax that will cause your relay to go into hibernation when you hit a certain limit. With a 1000GB/mo VPS plan you could set a conservative monthly limit of 900 GB (or a daily limit of 30 GB) and Tor will automatically stop serving well before you dip into extra costs.
The only downside to this is that your relay won't be very stable. For a certain amount of time every day/month your server will be idle, thereby limiting the amount of consistent help you can contribute. I've used the
RelayBandwidthRate setting to limit my bandwidth in an attempt to find a balance where I'm always relaying but never going over my limits. However, as throughput rates tend to fluctuate I'm not really sure if my settings are optimal.
Is there an easy way to determine what
RelayBandwidthBurst settings are optimal for a given monthly cap? Would traffic ever be consistent enough to to come up with a simple table that could be included in the setup guide?
Essentially I'm looking for something like this: