Cannot start kernel since there are no other nodes in the cluster to sync up for the database

If at least one of your nodes is online, see this article.

You should only perform these steps if all nodes have the error message in the title of this article.

WARNING: Performing these steps in the wrong sequence can lead to loss of your data.

Follow this 6-step process:

1. Stop of all of your nodes and make sure they're not looping trying to rejoin.
2. Create backups of Stardog home for all 3 nodes and transfer them off the nodes.
3. Use stardog-admin zk clear to clear ZooKeeper.
4. Start the most up-to-date node and allow it to join.
5. Once it joins, start each of the other nodes one at a time, allowing them to fully join before moving on to the next one.
6. Verify all of your data is there.

For general information on clusters, as well as instructions for installing and operating clusters, please see the docs.