I am trying to configure my Relay node in the best way possible to provide a stable connection to users. My provider offers a connection of 1 Gbit/second and 100 TBytes/month. Therefore, I have decided to use the following configuration to maintain an active connection every day with a certain limitation. Please let me know what you think and if there’s room for improvement. I have chosen to set a daily limit to avoid having to interrupt the connection for several consecutive days if the total monthly allowance is consumed much earlier than the end of the month.
# The network usage count resets every day at 0:00
AccountingStart day 0:00
# Daily maximum set to 33 GBytes
AccountingMax 33 GBytes
# Bandwidth set to 200 MBits
RelayBandwidthRate 200 MBits
If you have 100TB of bandwidth available monthly, you’d wand to set your AccountingMax to something much higher. A conservative value would be AccountingMax 1TBytes. This would allow up to 62TB/month (Up/Down combined) on months with 31 days with plenty left over.
Consider that 1 mbps = ~320GB data per month or 640GB bi-directionally. So with a RelaBandwidthRate of 200 MBits you may exhaust your AccountingMax before the end of the month (640GB * 200 = 128TB).
The biggest factor is can your node handle that amount of traffic? If it can you’ll likely have to tweak both parameters over a period of time if you really want to fully utilize all 100TB without having your node go offline due to accounting limits being hit.
I found this solution independently and it is easier than expected.
However, I noticed that a few other people had the same doubt as I did and so I created a document explaining how to calculate what I was asking for in the application and by doing so I was able to remove the statements AccountingStart day 0:00 and AccountingMax 33 GBytes since if the maximum availability offered was used I would still be able to stay under the limit made available by the provider.