Skip to main content

purgeMarketDataDatasets

Delete datasets: their v1 object, their v2 chunks, the rows that index them and their per-ref state (failure streak, invalid-ref entry). Irreversible. The filter must name an exchange (DATALAKE_PURGE_NEEDS_EXCHANGE). Needs org__datalake__manage on organizationId, a system organization, and a fresh purge-datalake step-up grant unless dryRun.

purgeMarketDataDatasets(
  organizationId: String!
  filter: MarketDataLakeActionFilter!
  dryRun: Boolean! = false
  expectedAffected: Int
  force: Boolean! = false
): MarketDataLakeActionResult!

Arguments​

purgeMarketDataDatasets.organizationId ● String! non-null scalar​

purgeMarketDataDatasets.filter ● MarketDataLakeActionFilter! non-null input​

purgeMarketDataDatasets.dryRun ● Boolean! non-null scalar​

purgeMarketDataDatasets.expectedAffected ● Int scalar​

On a real run, the count the operator's dry run showed. When the action would now affect a different number of rows it is refused (DATALAKE_ACTION_COUNT_CHANGED) and changes nothing.

purgeMarketDataDatasets.force ● Boolean! non-null scalar​

Purge datasets a running bot trades too. Without it such a purge is refused (DATALAKE_PURGE_DATASET_IN_USE) and changes nothing.

Type​

MarketDataLakeActionResult object​

What a lake admin action did, or with dryRun would do.