Tim Fox

We are currently using the pingcap parser. Apparently it was moved across 6 months ago

We should create a Jepsen https://github.com/jepsen-io/jepsen test suite for PranaDB

Something to consider. An alias is an indirection to a source and materialized view and the target can be changed after creation to switch the source to another place. E.g....

The Kafka rebalance protocol is not implemented in all Kafka clients (e.g. Segment IO). The protocol is designed, such that, when a rebalance is in progress, consumer apps are given...

It's getting too cluttered. In particular we should look at splitting out the encoding suff to other packages.

Prana currently uses hash partitioning to partition data into shards. Hash partitioning has an advantage when ingesting sequential data (e.g. for Kafka topics where the key is ascending) as data...

We should create a nice web UI which can be pointed at a Prana cluster and used to visualise various metrics (numbers of rows in tables, rate of ingest/egest etc)...

A lot of the time spent in SQLTests is due to deleting all table data that occurs when a source or MV is dropped. We could probably speed up SQLTests...

Dragonboat currently fsyncs on every raft commit. This can make latencies high, especially when running tests on a Macbook. For test runs, it would be nice if we could turn...