Data Recovery After Node goes down?

Say I want to build a ceramic data explorer(data indexer) for my app. I could access the live update of data throughout my application by subscribing to the relevant libp2p topics.

But what if my subscribing node goes down and recover? Is there a way for me to read historical messages before continuing to listening to the new updates? How does ceramic node itself handle down time and recovery? How to scale my node if the app got some traction?

1 Like

Not at the moment but this is something we are currently working on and planning to release for the coming ComposeDB beta.
Nodes will have the ability to index streams based on the anchor contract, so having a global visibility of streams present on the network.
The upcoming hard fork of Ceramic is necessary to support this feature, see HistorySync Hard Fork: Upgrade Your Node by Feb 15 for more information.


So for now, how do I find out all the live modesl and composites?

1 Like

if you want to see the list of all Models that have been published on the network, you can use the composedb CLI: composedb model:list. By default it outputs the models on Ceramic Mainnet, which won’t have any yet since we haven’t launched on mainnet yet, but if you set --network=testnet-clay you can see the Models that have been deployed to our Clay Testnet.

1 Like

lovely. Could I ask also how composites differ from old ceramic Models? Is there a meta-model concept?

Coooool, Can I ask how it works? It works by kh4q0ozorrgaq2mezktnrmdwleo1d stream?

if yes, is actually kh4q0ozorrgaq2mezktnrmdwleo1d a ‘fake’ stream, and it is only a DB table?

Is there an extra task that records all deployed models when any models commit in the corresponding network?

1 Like