Tim Fox

Currently sources are created and either fill from the whole topic or only receive latest messages. With a large topic you might want to fill from, say, the last 7...

When a new rows arrives on a source or MV often the previous value for the row needs to be loaded from storage. Performing a get for every incoming row...

When creating a new source against a topic with a lot of data, we should consider providing an out of band backfill to speed things up. This wold mean the...

We should allow a functional Prana cluster to be startable with just one node for development purposes.

When we pass through a RowBatch in a push DAG, an entry can be an addition, modification or delete - i.e. we pass prev row and current row. Instead we...

The current DDL mechanism can be improved simplified. We should * Maintain a table ddl_log in the system shard along with the tables table, this records the last ddl submitted...

For direct sends and consumes we should support the Kafka protocol directly so we can be accessed using standard Kafka clients

When working on k8s and tuning PranaDB it's very painful and slow to have to change config and redeploy the cluster each time. I would like to introduce a new...

We should allow for older data (cold-set) to be offloaded to S3 buckets (when running on AWS), this would allow Prana dataset to scale to virtually unlimited sizes. Blocks of...

We should create a plugin interface which allows protobuf definitions to be looked up from a schema registry. Right now, teams have to manually upload their protobuf definitions. With a...