Creating assemblies in UI
Create the components of an assembly using a guided wizard.
Set a database name and optionally choose to setup an external stream for data publishing into Insights via SDKs. The initial database has preset resources for each component of the database.
The first choice is to select an assembly which offers predetermined CPU, Memory, Size and Replica settings for databases, pipelines and streams; at this step, the name of the assembly is required
Configure your schema. Here the tables and columns that make up those tables can be added. Each table in your schema must include a timestamp column for partioning the data.
An assembly schema has a name and at least one table; a table has a set of properties and predefined kdb types. Each table schema must have a timestamp column assigned
Define the Table Properties for the selected schema, chief of which is the Partition Column, which must be a timestamp data column. The data column to use for real-time, interval and historic sort can also be set; typically, this is the timestamp column too.
Essential properties with partioning of time column. Required are definitions for Partition Column, Timestamp Column, Real-Time Sort, Interval Sort and Historical Sort. The latter three sorts all use the named time column from the table schema
Once the database and schema have been configured a summary of each will be presented. Users familiar with databases can look to customize database, but this isn't necessary for a starter assembly. An option to import data by creating a data pipeline is also available at this stage, but is not required.
Save & Deploythe assembly.
A completed assembly with pipeline details
A summary view of a completed assembly ready to be saved and deployed.
Associated pipelines should have a timestamp data column, and a designated partition column in the schema. Failure to do so could result in a reported error or an empty data set when exploring the data.
Should the assembly fail to deploy, check diagnostics for any reported errors.