Type Alias PbTable

pub type PbTable = Table;

Alias for Table.

Aliased Type

struct PbTable {

    pub id: u32,
    pub schema_id: u32,
    pub database_id: u32,
    pub name: String,
    pub columns: Vec<ColumnCatalog>,
    pub pk: Vec<ColumnOrder>,
    pub dependent_relations: Vec<u32>,
    pub table_type: i32,
    pub distribution_key: Vec<i32>,
    pub stream_key: Vec<i32>,
    pub append_only: bool,
    pub owner: u32,
    pub fragment_id: u32,
    pub vnode_col_index: Option<u32>,
    pub row_id_index: Option<u32>,
    pub value_indices: Vec<i32>,
    pub definition: String,
    pub handle_pk_conflict_behavior: i32,
    pub read_prefix_len_hint: u32,
    pub watermark_indices: Vec<i32>,
    pub dist_key_in_pk: Vec<i32>,
    pub dml_fragment_id: Option<u32>,
    pub cardinality: Option<Cardinality>,
    pub initialized_at_epoch: Option<u64>,
    pub created_at_epoch: Option<u64>,
    pub cleaned_by_watermark: bool,
    pub stream_job_status: i32,
    pub create_type: i32,
    pub description: Option<String>,
    pub incoming_sinks: Vec<u32>,
    pub initialized_at_cluster_version: Option<String>,
    pub created_at_cluster_version: Option<String>,
    pub retention_seconds: Option<u32>,
    pub version_column_index: Option<u32>,
    pub cdc_table_id: Option<String>,
    pub maybe_vnode_count: Option<u32>,
    pub version: Option<TableVersion>,
    pub optional_associated_source_id: Option<OptionalAssociatedSourceId>,
}

Fields

id: u32

schema_id: u32

database_id: u32

name: String

columns: Vec<ColumnCatalog>

pk: Vec<ColumnOrder>

dependent_relations: Vec<u32>

For cdc table created from a cdc source, here records the source id.

TODO(rc): deprecate this by passing dependencies via Request message

table_type: i32

distribution_key: Vec<i32>

stream_key: Vec<i32>

pk_indices of the corresponding materialize operator’s output.

append_only: bool

owner: u32

fragment_id: u32

vnode_col_index: Option<u32>

an optional column index which is the vnode of each row computed by the table’s consistent hash distribution

row_id_index: Option<u32>

An optional column index of row id. If the primary key is specified by users, this will be None.

value_indices: Vec<i32>

The column indices which are stored in the state store’s value with row-encoding. Currently is not supported yet and expected to be \[0..columns.len()\].

definition: String

handle_pk_conflict_behavior: i32

Used to control whether handling pk conflict for incoming data.

read_prefix_len_hint: u32

Anticipated read prefix pattern (number of fields) for the table, which can be utilized for implementing the table’s bloom filter or other storage optimization techniques.

watermark_indices: Vec<i32>

dist_key_in_pk: Vec<i32>

dml_fragment_id: Option<u32>

A dml fragment id corresponds to the table, used to decide where the dml statement is executed.

cardinality: Option<Cardinality>

The range of row count of the table. This field is not always present due to backward compatibility. Use Cardinality::unknown in this case.

initialized_at_epoch: Option<u64>

created_at_epoch: Option<u64>

cleaned_by_watermark: bool

This field is introduced in v1.2.0. It is used to indicate whether the table should use watermark_cache to avoid state cleaning as a performance optimization. In older versions we can just initialize without it.

stream_job_status: i32

Used to filter created / creating tables in meta.

create_type: i32

description: Option<String>

This field is used to store the description set by the comment on clause.

incoming_sinks: Vec<u32>

This field is used to mark the the sink into this table.

initialized_at_cluster_version: Option<String>

Cluster version (tracked by git commit) when initialized/created

created_at_cluster_version: Option<String>

retention_seconds: Option<u32>

TTL of the record in the table, to ensure the consistency with other tables in the streaming plan, it only applies to append-only tables.

version_column_index: Option<u32>

This field specifies the index of the column set in the “with version column” within all the columns. It is used for filtering during “on conflict” operations.

cdc_table_id: Option<String>

The unique identifier of the upstream table if it is a CDC table. It will be used in auto schema change to get the Table which mapped to the upstream table.

maybe_vnode_count: Option<u32>

Total vnode count of the table.

Use VnodeCountCompat::vnode_count to access it.

• Can be unset if the table is created in older versions where variable vnode count is not supported, in which case a default value of 256 (or 1 for singleton) should be used.
• Can be placeholder value Some(0) if the catalog is generated by the frontend and the corresponding job is still in Creating status, in which case calling vnode_count will panic.

Please note that this field is not intended to describe the expected vnode count for a streaming job. Instead, refer to stream_plan.StreamFragmentGraph.max_parallelism.

version: Option<TableVersion>

Per-table catalog version, used by schema change. None for internal tables and tests. Not to be confused with the global catalog version for notification service.

optional_associated_source_id: Option<OptionalAssociatedSourceId>