Jack Ye
e9774ed168
Expose `publicationIntervalRows` on TypeScript refresh options and
forward it as `publication_interval_rows`. Accept positive safe integers
and preserve whole-fragment publication when omitted.
```typescript
const job = await table.refreshColumnAsync("embedding", {
publicationIntervalRows: 10000,
});
await job.wait();
```
Document the threshold semantics and add coverage for serialization,
omission, invalid values, safe-integer boundaries, and Embedded
rejection.
Closes ENT-3755.
2.5 KiB
@lancedb/lancedb • Docs
@lancedb/lancedb / RefreshColumnOptions
Interface: RefreshColumnOptions
Options for Table.refreshColumn and Table.refreshColumnAsync. Every
option is LanceDB Enterprise only; a LanceDB Embedded refresh ignores
batchSize, gpuClass and gpuCount and rejects the rest, including
resourceBudget.
Properties
batchSize?
optional batchSize: number;
Maximum rows per Function call; must be greater than zero.
errorPolicy?
optional errorPolicy: string;
"fail" (default) stops on a row error; "skip" writes null and
records the error, subject to the failure limits.
gpuClass?
optional gpuClass: string;
GPU class for Function workers.
gpuCount?
optional gpuCount: number;
Positive aggregate GPU budget for Function workers.
initTimeoutSeconds?
optional initTimeoutSeconds: number;
Deadline for Function initialization, in seconds.
maxConcurrency?
optional maxConcurrency: number;
Override the registered Function's maximum concurrent user-code calls for this logical execution.
maxExecutorProcessesPerWorker?
optional maxExecutorProcessesPerWorker: number;
Cap on executor processes per Function worker.
maxFailedFraction?
optional maxFailedFraction: number;
Under a skip policy, fail the job once more than this share of rows (0 to 1) has failed.
maxFailedRows?
optional maxFailedRows: number;
Under a skip policy, fail the job once more than this many rows have failed.
numWorkers?
optional numWorkers: number;
Worker count for an expression column; Function columns reject it.
publicationIntervalRows?
optional publicationIntervalRows: number;
Row-count threshold for publishing completed Function split ranges. Must be a positive safe integer. Omit for whole-fragment publication. Actual publications can be larger; remaining rows publish at completion. Requires LanceDB Enterprise with non-Blob Function outputs.
resourceBudget?
optional resourceBudget: FunctionWorkerResourceBudget;
Complete aggregate CPU and memory budget for a registered Function
refresh. gpuCount inside the budget defaults to zero.
rowTimeoutSeconds?
optional rowTimeoutSeconds: number;
Deadline for one Function call, in seconds.
where?
optional where: string;
Only refresh the rows matching this SQL filter.