INC_04030373: IO error while reading snapshot file...
I have noticed that if I have a schema in an incremental load and have to stop/start the loader service, this error occasionally occurs. Is there a simple way to clear this error from the server-side? My workaround is to initiate a full load on the offending schema and then another incremental load. This seems to clear the .zxc file.
Any insight is welcome.
Part of the schema refresh job, the snapshots will be written. If the loader service is force stop during the load, the snapshot file may be missing or corrupted. To recreate these files, you can run "Load from Staging". You don't have to run a full load.
The difference between "Load from staging" and "Full load" is that the extraction and transformation phase will be skipped in load from staging. The parquet files generated from prior extraction will be used to load to Incorta engine and the joins, the formula columns, the incorta derived tables will be recalculated, and the generated DDM files (snapshot files) will be written and synchronized to analytics services.
Stop loader services and stop loading job may cause the issue, so we should avoid stop loader service by killing it from backend, clicking the stop multiple times from CMC, or running the stopService.sh CLI multiple times from command line in a production system.