[Epic] Extended query protocol

[scsmithr]

https://www.postgresql.org/docs/current/protocol-flow.html#PROTOCOL-FLOW-EXT-QUERY

benchbase

Benchbase uses prepared statements, so attempting to get the tpch benchmark running would probably be a good test case.

$ git clone --depth 1 https://github.com/glaredb/benchbase.git
$ cd benchbase
$ git checkout glaredb
$ ./mvnw clean package -P postgres
$ cd target
$ tar xvzf benchbase-postgres.tgz
$ cd benchbase-postgres
$ java -jar benchbase.jar -b tpch -c config/glaredb/sample_tpch_config.xml --create=true --load=true --execute=true

features

• Parse Support placeholders when planning queries #122
• Bind Support placeholders when planning queries #122
• pgsrv: Respect max rows paramter when executing a statement #132
  • Run query planned from portal
• Describe
  • portal
    • Ddl feat(pgsrv): Implement start of postgres extended query protocol #117
    • Write
    • Query feat(pgsrv): Implement start of postgres extended query protocol #117
    • Transaction
    • Runtime
  • pgsrv: Describe prepared statement #131
• Sync feat(pgsrv): Implement start of postgres extended query protocol #117
• Terminate feat(pgsrv): Implement start of postgres extended query protocol #117

[greyscaled]

This epic needs re-work and break-out. We don't need everything from the extended query protocol, but there are use-cases where someone will need to run a prepared query.

[greyscaled]

Lower until we get to a certain milestone in product of supporting analytical use cases.

[greyscaled]

We have stories for the prepared statements. Other than those, does an analytics db (non-transactional) need to implement more from the extended query protocol?

I know we have a long-term vision of HTAP, but if it's far enough away from our immediate use case, I'd rather close the ticket since it's stale and re-open an epic for that when we're working towards that goal more actively.

[greyscaled]

Stale - can reopen later