Cloudstroller.xyz: local cloud storage / backup service.

newby_investor

Executive Member
Joined
Aug 8, 2018
Messages
5,195
Reaction score
3,623
Location
Cape Town
So I'm busy developing the capability to host a cloud storage / backup server aimed at individuals / families and small businesses.

My main selling points would be the following:
  • Locally hosted (pay in rands, faster upload / download speeds since it's not an international connection)
  • Privacy (you're the customer, not the product)
  • No ads / datamining / anything like that
  • Completely open-source, open-standard software
  • Assistance with disaster recovery (i.e. theft, ransomware, hard drive failure, accidental deletion, etc)
I have a limited beta going on at the moment and it seems to be working well. I am planning within the coming couple of months to start offering a slightly more open beta, perhaps with a discount to early adopters, before full launch.

I personally think there's a great need for off-site backups for most people who do anything on a computer, though admittedly many people don't until they have their first theft / hard drive crash / etc.

I'd like to get a feeler as to what the needs are in the local tech community. Do you have family photos that you'd like to keep safe? Do you run a small business whose operations would be severely compromised if your laptops got stolen? How much do you need backed up, and how much do you think would be a reasonable cost to pay for the service? What other services do you think would be useful?

I have some ideas in mind already around pricing, support, sizes, etc. but I'd like to get a bit of a feel from others.
 
It will come down to pricing (which its unlikely you will be able to compete in). Individuals, families and SMBs are likely to use free cloud storage products already on the market and integrated with their mobile devices.
 
It will come down to pricing (which its unlikely you will be able to compete in). Individuals, families and SMBs are likely to use free cloud storage products already on the market and integrated with their mobile devices.
The pricing I had in mind is actually fairly competitive with all but one of the existing international providers. Sure, not everyone will be willing to pay for services when there's free stuff available. But I think there are at least some who are willing to pay for a better service, especially if it's one that respects your privacy and offers more local support than provided by the international competitors.
 
The pricing I had in mind is actually fairly competitive with all but one of the existing international providers. Sure, not everyone will be willing to pay for services when there's free stuff available. But I think there are at least some who are willing to pay for a better service, especially if it's one that respects your privacy and offers more local support than provided by the international competitors.
I like the idea, but you are competing against $10 a terabyte offers.

Regarding the privacy matter, thats real, and the value added service you offer is good. If you can sell that to people then go for it. While you can offer different levels of privacy and service, what about security?
 
Might be interested depending on price /performance and the guarantee that you'll still be here in 10 years.
 
I hosted stuff for clients for years; started with a single server at home until i had a full size rack with multiple servers/ups's/etc.

Unfortunately over time the cost of hosting (electricity/insurance/etc) became less lucrative + with all the free/inexpensive services out there it became harder to compete.
You would need a very large client base to make decent profit which comes with a fair amount of customer support (yet another cost)

As I was doing it on the side the time/effort/small profit just wasn't worth it anymore.
Rather get the clients to pay for some online storage/service and then help them manage it.
 
Just in time for the weekend, I am ready to start properly beta-testing my offering. I am starting with NextCloud, in coming weeks I am looking to get a SeaFile server ready. My intention is to offer both, according to clients' preferences.

If anyone is interested and has a bit of time to load my server, mail me and I will hook you up with a free 1GB account.

I will also offer proper amounts of space (500GB, 1TB) for a discounted price, which will be locked in for beta testers as a way of saying thanks for your help. Mail for more details!

[email protected]
 
What DC is the server located in?
Backups/redundancy?
 
What DC is the server located in?
Backups/redundancy?
Not completely ready to disclose all my info yet, but location is in Cape Town, and there is local redundancy, but with no off-site backups as yet - still in beta-testing phase.

Mail me for more info if you are interested.
 
Not completely ready to disclose all my info yet, but location is in Cape Town, and there is local redundancy, but with no off-site backups as yet - still in beta-testing phase.

Mail me for more info if you are interested.

Locally hosted isn’t much of an advantage anymore due to peering and local presence of many cloud providers. There might be room for a fully encrypted niche solution with physical media shipping of data in case of emergency etc.
 
Locally hosted isn’t much of an advantage anymore due to peering and local presence of many cloud providers. There might be room for a fully encrypted niche solution with physical media shipping of data in case of emergency etc.

Good point about local peering, but I like to think that local support is still an advantage :-)

Physical media shipping is definitely a feature I intend to implement - I've tried the process on my own systems and it works.

Over the weekend I have gotten my Seafile server properly configured - Seafile supports client-side encryption, however Nextcloud does not at this stage. That being said, I do encourage and support the use of encryption through third-party such as Cryptomator which can be used with any cloud-storage provider. I prefer to leave this choice up to the user.

The main distinguishing feature that I am hoping to offer is, if you don't really feel like being continuously data-mined by large cloud businesses which shall not at this point be named, that there is a small, reasonably priced independent option. I make my money from you, the customer, not by using your data to sell on to advertisers or others.
 
My suggestion would be to look at alternative providers like Dropbox and Mega.
Dropbox gives Business customers unlimited disk space on some of their plans.
Mega is similar, except storage is limited in comparison to DB

If you could look at the pros and cons of other providers, you could build a better system.
 
The real question is, can you compete with Backblaze?

I am considering moving to them however, if your offering is better then there will be need for thought
 
The real question is, can you compete with Backblaze?

I am considering moving to them however, if your offering is better then there will be need for thought
Backblaze is in some ways a model for what we'd like to do. They've built a good business. And from what I know, they started off similar to what we have - a server room in an office somewhere. And grew from there.

We are starting out offering essentially an open-source solution more similar to Dropbox than Backblaze. The client gets the choice of either Seafile or Nextcloud. The next thing we'll look at is possibly something similar to Backblaze's $5/m whole-computer backup, but their software is proprietary, and we prefer to stick to open-source only. Similar packages do exist, we are just not confident enough in them to recommend them yet, as we don't want to create a frustrating experience for a client. We have a small dev team currently investigating options.

And ultimately, we like to think that the internet shouldn't be a monoculture where there's one and only one way of doing things. The user should feel as though he has the choice to go to any provider and get a decent customer experience.
 
So can you tell me what you intend to offer or some indication of pricing and structure?

Thanks
 
Backblaze is in some ways a model for what we'd like to do. They've built a good business. And from what I know, they started off similar to what we have - a server room in an office somewhere. And grew from there.

We are starting out offering essentially an open-source solution more similar to Dropbox than Backblaze. The client gets the choice of either Seafile or Nextcloud. The next thing we'll look at is possibly something similar to Backblaze's $5/m whole-computer backup, but their software is proprietary, and we prefer to stick to open-source only. Similar packages do exist, we are just not confident enough in them to recommend them yet, as we don't want to create a frustrating experience for a client. We have a small dev team currently investigating options.

And ultimately, we like to think that the internet shouldn't be a monoculture where there's one and only one way of doing things. The user should feel as though he has the choice to go to any provider and get a decent customer experience.
I've got both Nextcloud and Seafile.
Will you customise these software solutions, or will you only be a 'host' of the standard offerings?
 
I've got both Nextcloud and Seafile.
Will you customise these software solutions, or will you only be a 'host' of the standard offerings?
At the moment, we're only hosting. A custom solution may be eventually in the pipeline, but as these are both AGPL, any customisation that we do would need to be released as source code, so we are more likely to submit any changes or improvements that we make back upstream than to try and distinguish ourselves. It also depends on demand.

The decision to start with these two was made basically to assess the market, and make something that's easy for the average tech-savvy consumer to start off with, as they're quite similar i.t.o. usability to the other things that people use on a day-to-day basis.

The next step, rather than offering customised Nextcloud- or Seafile-type services, would be specifically a backup product - i.e. something like Backblaze's product, which is not intended for syncing or filesharing, but more disaster recovery (i.e. against theft, hard drive crash or ransomware). The idea would be to offer this at a lower price per TB than the standard cloud storage option. This will be feasible as users would be writing periodically but only reading occasionally. As I mentioned, this is what our dev team is currently looking into. We'll be testing amongst ourselves quite thoroughly before we make it available for customers.

(Disaster recovery can currently be done using Seafile or Nextcloud, but if that's all you need then these have a lot more functionality that isn't really going to help.)

So can you tell me what you intend to offer or some indication of pricing and structure?

Thanks

What we intend to offer I summarised above, we're still very much in the beta phase though. We hope to offer a stable offering, but we do realise that what the consumer wants may not be quite what we think the consumer needs. So we're starting off simply hosting Nextcloud and Seafile services :)

We intend to keep the pricing very simple for the time being, at around R60/TB/m. We may do a bit of a sliding scale, e.g. R35/500GB/m, R110/2TB/m (i.e. buy bulk and save). We also do plan to offer a discount for quarterly, semi-annual or annual payments, though that will likely come only after we've satisfied ourselves that we're ready to go out of beta.

As I noted previously, during the beta we will be offering a discounted price which will be locked in for clients who stay with us, to show our appreciation.
 
Do you have any AUP's on what users are allowed to store on your servers?

I'd like to dig deeper into your infrastructure if I may:
  • Are you hosting in a DC? If so, are you running your own hardware or is it leased?
  • What is the SLA on your environmental variables (HVAC, electricity, etc)
  • What is your actual raw capacity? If I and 5 of my friends sign up and we all upload 10TB are we going to saturate your servers?
  • What are your plans to scale out/up when you start running out of space in your current environment?
  • Connection wise, are you running 100Mbps/1Gbps/10Gbps? Do you have redundancy if one link goes down?
I think these are my main pain points to determine what your reliability will be.
 
Do you have any AUP's on what users are allowed to store on your servers?

I'd like to dig deeper into your infrastructure if I may:
  • Are you hosting in a DC? If so, are you running your own hardware or is it leased?
  • What is the SLA on your environmental variables (HVAC, electricity, etc)
  • What is your actual raw capacity? If I and 5 of my friends sign up and we all upload 10TB are we going to saturate your servers?
  • What are your plans to scale out/up when you start running out of space in your current environment?
  • Connection wise, are you running 100Mbps/1Gbps/10Gbps? Do you have redundancy if one link goes down?
I think these are my main pain points to determine what your reliability will be.
Regarding AUPs - at this stage, no. Our draft T&C document states only that users are responsible for what they store, and that we will comply with any requests from law enforcement for data.

I don't want to give away too many details about our business at the moment. Up to now, while we have been providing off-site backup and recovery services for several existing clients on an ad-hoc basis, it's not our core business. We are currently looking to expand into this space, as it's something we're quite passionate about.

I'll share the following details though. We are not currently in a datacentre, but a server room at our office premises. High availability isn't guaranteed, I'll admit. We have sufficient ventilation, and there are UPSes to protect the equipment, but not sufficient keep going during 2 hours of load-shedding, for instance.

We are running our own hardware, with a modest capacity (Order tens of TBs). Current clients use far less than we'd originally anticipated, around 5% of our capacity. If we get greedy and sign up too many clients, we could be saturated possibly within a few days. Our plan is to accept clients until space sold (even if the client doesn't actually use the space) is roughly 75-80% of capacity available, stop accepting clients, test for a bit, then double the capacity and repeat. We have capital available to scale out to about 100 TB at the moment (physical rack space for considerably more if the cash flows), but this is largely dependent on how enthusiastic the uptake of the offering is. Unfortunately, we have no link redundancy at the moment, as only one fibre provider covers our premesis. Our ISP advises that our maximum (symmetric) link speed is 300 Mb/s, and while we're not using that at the moment (we haven't had a need up to now) it takes a matter of hours for them to increase it.

If the uptake is enthusiastic, we will almost certainly move into a datacenter, though this will need to be planned carefully. At the moment I think we have the capacity to serve a fair number of clients quite well, and to add value by providing support - this has up to now been closer to our core business, not just providing an online-only service but a human being to advise and assist you in managing your data.
 
Top
Sign up to the MyBroadband newsletter
X