zenstore

zenstore is a simple JSON storage backed by leveldb, node-levelup. You can easily post or put data to the storage and receive it by a simple get.

It is designed for simple use. Just request a new store by visiting the web-frontend or by calling the /create route.

What it will become in the future

zenstore is supposed to be a simple "back-end" like structure for the "Internet of Thungs" (IoT) - IoT-Wikipedia, IoT-McKinsey, IoT-Europe. All this is very cool, but how to connect and store the data of things?

zenstore tries to solve the problem, it gives everyone the possibility to just store and exchange data! This is done via an easy to use GET/PUT/POST (REST) API as well as a TCP service (for the high end users).

BTW: Do you want to use zenstore from the CLI? Go to zenstorecli

Link 'secret' storages, that's like naming it

This one is quite cool, if you want to give somebody an easy read access just use the createZenlink API call. With this goodie it is possible to hide the random generated 'secret' zenstorage ID behind a human readable name. E.g. your zenlink ID is c739a87f21d1740fbf83411d3132043780043a8b you can link it by calling http://[...]/createZenlink/c739a87f21d1740fbf83411d3132043780043a8b/yourbeautifulname. Now you can share this zenstorage with ease: http://[...]/zenlink/yourbeautifulname.

Linking Computations

This feature is quite cool - well, personally I like it... - if you don't want to share your data rawly, you can link a computation with your data set. Look at the example:

$ curl http://localhost:1337/67030dcf8eb3a25cc14dc223ccbb5234b786a90a
{"test":3}

$ curl \
> -d '{"script": "var output = {\"mynum\": data.test + 2}"}' \
> http://localhost:1337/linkComputation/67030dcf8eb3a25cc14dc223ccbb5234b786a90a
{"linkComputation":{"id":"67030dcf8eb3a25cc14dc223ccbb5234b786a90a","script":"var output = {\"mynum\": data.test + 2}"}}

$ curl http://localhost:1337/67030dcf8eb3a25cc14dc223ccbb5234b786a90a
{"mynum":5}

$ curl http://localhost:1337/unlinkComputation/67030dcf8eb3a25cc14dc223ccbb5234b786a90a
{"unlinkComputation":"67030dcf8eb3a25cc14dc223ccbb5234b786a90a"}

$ curl http://localhost:1337/67030dcf8eb3a25cc14dc223ccbb5234b786a90a
{"test":3}

The computation is done on the server, directly where your data is. This means, it is possible to share your results filtered, aggreated or even completly changed. BTW it does work with linked storages as well:

$ curl http://localhost:1337/zenlink/abc
{"test":3}

$ curl \
> -H "Content-Type: application/json" \
> -d '{"script": "var output = {\"mynum\": data.test + 2}"}' \
http://localhost:1337/linkComputation/67030dcf8eb3a25cc14dc223ccbb5234b786a90a

$ curl curl http://localhost:1337/zenlink/abc
{"mynum":5}

Quite cool, isn't it. And it should be save as well because we are using the node.js vm module and that runs everything you pass in inside a sandbox. If you receiver the raw data after you linked a computation your computation fails.

To make it work you always have to follow a simple structure: var output = ... is the variable returned to the request. data.<yourdata> holds the stored data and {"script": } is the anchor for a computation. From then, it is pure JS & Node love.

LICENSE

MIT

Roadmap

• implement search and atomic operations for one zenstorage
• implement preview for data-types at the "follow" subpage
• implement basic-auth (maybe, atm. I do like the sha1 secret-ids)