- Preparing search index...
- The search index is not available
When the SDK writes an updated checkpoint event ID back to the RStreams bus, it checks if someone has changed the checpoint out from underneath the bot. In practice, this would only happen in one of the following two scenarios. 1) a developer is running independent sets of code as though it were the same bot erroneously or 2) a type of fanout is happening where the multiple copies of the same code is running as the same bot. In scenario 1, this flag will be false and the developer will get an error back. In scenario 2, this flag will intentionally be set to true by whomever is writing the fanout code because it's OK that someone else changed the checkpoint since the last time it was read by some code.
This is only used in advanced scenarios.
Checkpoint after this number of records (events).
Checkpoint after this amount of time expires.
Note, this type is any one of the valid durations the Moment JS library can take: Duration | number | string | FromTo | DurationInputObject.
Generated using TypeDoc
These options allow a developer to tell the SDK how often it should automatically checkpoint. To checkpoint is to make a call back to the RStreams bus instance and tell it that a given bot has successfully read up to the event ID included in the API call. This ensures that the next time a bot needs to run, the last known checkpoint position is remembered by the bus and the SDK can begin reading from that point in the queue forward in time.
It is used by adding an instance of this as a step in the pipeline itself to determine when to checkpoint.
The SDK will checkpoint as soon as either the
recordsortimeconstraint is met.