[BLOG] The Electric Car Example Applied to Decentralized Cloud Storage

Building on the Storj / Tardigrade platform. Decentralized apps, information about getting started, learning about the Open Source Partners program, and more


Satellite failover will go a long way in taking away some of the centralization worries. It’s good to see that’s on the roadmap. Another option for larger companies could be if they could host their own metadata or a copy of it. That will still require quite a bit of storage and coordination with tardigrade satellites, but could significantly increase availability of data.

1 Like

While I understand the current need for the satellite to be like it is, I think that advertising Tardigrade as “decentralized and therefore more reliable than competitors” is a bit misleading. Well, at least I do not think that Amazon and others just have one server or a few servers in one datacenter and without backups, so, however many bombs are needed to take out S3, the same amount of bombs will most likely take out the satellite as well.

Is the Tardigrade satellite more reliable than a singe server in some datacenter (or a few redundant servers in a single datacenter)? Yes. Is the satellite more reliable than Amazon? I don’t know.

The way I understand it cloud storage on the major cloud platforms are set up in a fairly similar way, with servers responsible for knowing which data is where and others that actually store the data. I’m sure the former would have similar risk, but are also much easier to make redundant as there is less storage involved. But I think Storj would be more reliable for the data storage part of the equation.

Also note that the word reliability is used. I think taking out a satellite is mostly a risk for availability and not so much reliability. So to that extent I think the claim can be made.

Which means that Storj is not that different.

Without metadata the client will not be able to access his files. If the metadata is lost permanently, so are the files.