Time Series Data Storage
The Time Series block stores data on disk only if the publish to Predix Time Series fails or data comes in too quickly to store in memory.
Store on Failure
uses in-memory queueing and multi-threaded publishing to increase the throughput of the application in situations where data is accumulating faster than the max batch interval
timer. Data is only stored persistently if the publish request to Time Series fails or if input data is received much more quickly than it can be sent.
For information on the policy
field in the Cloud Gateway configuration file, see Time Series Publisher Block Config.