absurdfarce
commented
3 years ago Note that https://github.com/datastax/adelphi/issues/107 is adding support for a large collection of data types directly, enough to cover commonly used types in schemas. We still need to cover some additional cases:
- A few types still aren't covered there... consider inetaddress
- Support for the collection types (list, map, set)
- Probably need a good story around the handling (or lack thereof) of UDTs
Feature Request
Description
https://github.com/datastax/adelphi/issues/86 implemented a base framework for generating nosqlbench configs from an input schema. To ease the implementation we only added support for "text" types; this should be expanded to support as many CQL types as we reasonably can.
Note that this requires defining nosqlbench distributions for values for each supported type, so there may be a fair amount of twiddling there in order to get things like inetaddress working.
Describe the solution you'd like
As of this writing the impl has a fairly straightforward method for defining nosqlbench distributions/sequences in code (see https://github.com/datastax/adelphi/pull/96/files#diff-b50d6211d55bd7434d67c456c88581087f9c2c61e279c59891fa474667fc3485R8-R11). Ideally any impl here would just piggy-back off of that work and expand it somewhat.
Note that the list of types we can expect to see here will have to be extracted from the Python Cassandra driver
Based on the progress completed in #107 the following types are expected via this issue:
┆Issue is synchronized with this Jira Task by Unito ┆Issue Number: AD-44