Struct InternalTopicCheckpointStore 

pub struct InternalTopicCheckpointStore { /* private fields */ }

A [CheckpointStore] backed by a compacted internal Kafka topic on the target cluster.

Each flow gets its own key, flow_name, within the shared compacted internal state topic. On restart, load fetches the last value for that key and recovers the exact partition offsets that the worker had reached.

Implementations

impl InternalTopicCheckpointStore

pub async fn start( target_bootstrap: &str, flow_name: &str, security: Option<ClientSecurity>, ) -> Result<Self, ConnectError>

Ensure the compacted offset topic exists on the target cluster, build a producer, and return a store keyed by flow_name.

Errors

Returns [ConnectError::Offset] if the topic cannot be created or the producer cannot connect.

pub async fn start_with_policy( target_bootstrap: &str, flow_name: &str, security: Option<ClientSecurity>, client_resource_policy: ClientResourcePolicy, ) -> Result<Self, ConnectError>

Start with the deployment’s client resource policy.

Errors

Returns [ConnectError::Offset] if the topic cannot be created or the producer cannot connect.

Trait Implementations

impl CheckpointStore for InternalTopicCheckpointStore

fn save<'life0, 'life1, 'async_trait>( &'life0 self, offset: &'life1 SourceOffset, ) -> Pin<Box<dyn Future<Output = Result<(), ConnectError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Persist offset as the latest committed position. Called only after the records it covers are durable in the sink.

fn load<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<Option<SourceOffset>, ConnectError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Load the last persisted position, or None if the store never saved one, which is a fresh start. The runtime calls this once before the first poll.