IOTA met with Couchbase, opening a series of discussions at the table; In the discussion of the permanodos is one with a long history and is one that always seems to return. With the publication of local snapshots it becomes increasingly uncertain if your data will remain alive for more than 30 days in the tangle. However, with the local snapshots it is also possible to separate the worries! And this is really great!
For our project, as for many other projects; We want to choose if we keep the data available in the tangle or not for specific purposes of the application. Snapshotting is simply put-a headache-when it comes to this and with activated localsnapshotting it becomes increasingly difficult to keep the data alive in the tangle. Ideally we want to selectively save the data and be able to do what we do not want to store, but for that we must first be able to store everything and move from there.
Then follow us on Google News!
So I decided to do a little experiment to implement a persistence provider different from RocksDB and ZeroMQ. One for an external database , couchbase!
The benefits of Couchbase
By using an external persistence provider, we can keep the relatively small IRI instance running without sacrificing data storage and availability capabilities.
- We can run multiple IRI nodes that use back-end clustering so that we can run it with an update close to 100%.
- We can configure IRI nodes with different roles: I activate most of the live tangle and some IRI nodes exclusively to access the historical parts of the tangle.
- The API has not changed!
- We can independently scale the back-end storage service without ever having to turn off our IRI nodes.
It opens the search for semi-permanents that can store data selectively.
A few months ago, during a great effort of the community to create a permanode (localsnapshotting etc was all unknown at that time) I wrote a document with some considerations that are still maintained in this discussion (Wide-column, Graph, Key value etc. CHAP)
The conclusion was to choose for RiakDB with a personal bias to have a better experience compared to couchbase; and the most advanced features that are part of the open source version (such as the replication of cross-data centers). Members of the community reported that Riak de Basho’s team was in trouble with a great possibility that Riak was discontinued.