You can host a NuBot instance anywhere that you can run Java. A static IP is not a necessity (as long as the API request/response loop can reach the bot). While testing I’ve run NuBot locally (on OSX and Linux, but it also works on Windows), within a VM on my desktop, and remotely on AWS, DigitalOcean, and Azure (it will also work on other VPS).
I know that @KTm uses a cloud service to host at least part of her infrastruture, so it’s already “battle tested” using remote servers. As long as you employ good security practices, and maintain backups, you should be fine.
This is a very hard question to accurately answer, it really will depend on how you structure your operations. If you are offering an NBT/fiat pair as a liquidity provider (and aren’t offering dividends or have other operational needs that can’t currently be automated) it won’t be more than a few (2 or 3) hours a week to review how the operations are going and to adjust as needed.
NuBot and its related services are also getting easier to operate every day as the set of tools available to a custodian get better and better.
There’s no reasonable way for the team to offer a warranty on performance, so any losses incurred would be the responsibility of the custodian. If you are operating on a NBT/USD pair, it should be exceedingly difficult to lose anything due to bot malfunction if you configure it correctly. Any time you move to an environment where exchange rates come into play (NBT/crypto, NBT/non-usd-fiat), you introduce variables.
I don’t remember offhand if NuBot sends an email when it shuts down and I know it isn’t always possible to send a message if it was (e.g., if the computer it is running on goes down, the bot won’t “know” and can’t send a message), but it does have an emailer function so it is possible. External steps can be taken to set up services that will automatically restart the bot if certain strings of text are discovered in the log files, so automation is definitely possible.
Manual steps that may be required for a custodian’s operations:
- Setting up the exchange account(s) that NuBot (or NuBots) is/are connected to,
- Setting up NuBot on the hardware you intend to run it on,
- Other automation or configuration development that you want to do other than the default configuration,
- Reporting on operations (if it’s part of your agreement with shareholders). This is getting easier with the tools we are developing, but in some cases, you may prefer to manually prepare the reports.
- Dividend distribution (acquiring the peercoins, processing the distribution, etc.).
We need to make it more convenient, to a point. The tools and services to support a custodian can always be improved and we’re spending a lot of time each week doing exactly this. What we don’t necessarily want to do is to make it “stupidly easy” – every custodian in operation is a benefit and a risk to the Nu network. We need to balance the needs of NBT and NSR holders, custodians, and the larger community to best ensure due diligence on all sides.
At the most fundamental level, that is all that you need to do. NuBot doesn’t need to be compiled from source (there are binaries available) and it doesn’t actually even need to be installed (it can just be launched from the command line after customizing the options.json configuration file).
I consider the large compensation requests that we’re seeing to be a by-product of this being a new technology. This is a period of price discovery for prospective custodians. They are taking risks and want to make sure that they are adequately compensated for those risks. Until more LPCs have actually served it’s hard to point to a rate and say “that’s too high” or “that’s not high enough for the risk” because we really don’t know. The structure of how a custodian operates is entirely up to them.
For example: A LPC operating on a NBT/USD market could bring $10k in USD to the table, buy 5k NBT from an existing market, and then start to offer dual-side liquidity with a spread of $0.01 over the exchange commission if they so chose to. They don’t need to ask for compensation to do this, the risk is effectively limited to exchange default, and their profits would be generated from the spread.
Now, is that worth it to a custodian to do that? Maybe not for $10k worth of liquidity, but it may be if the base liquidity is higher with a corresponding increasing in volume. If you trust the exchange, the risks may be even lower, so for that custodian bringing $10k in liquidity it could be that they are happy requesting 2.5% to 10% in NSR or NBT to provide the service – the more LPCs vying for compensation, the more competitive the bids will become until a market develops.
