Batch ingest to druid

Because of the task failure, only part of the tasks in a certain time range of Druid are transformed into segments. Later, when we want to repair this part of the data by batch import, we find that the data has indeed been imported, but the original segment has not been replaced and still exists. What is the reason for this and what is the solution?

Please check if the following documentation helps:

http://druid.io/docs/latest/ingestion/update-existing-data.html

you have to do re-indexing of the data.

Thanks & Rgds

Venkat

When I use stream tranquility to write data to druid, my shardsNum is set to 3, so three real_time tasks are generated at each time point. while these three tasks are executing, two of them successfully generated two segments and closed automaticlly . However, the rest of the task still hanged several hours and not closed and there is no segement generated and no error logs. i have to killed this task. at last I tried to fix these three segments in this time span by Hadoop batch index task from hdfs but found that three segments generated by the ingested data could not replace the first two stream tranqulity segments . what is the pb and Is there any solution? Thanks.

在 2019年2月12日星期二 UTC+8下午6:34:17,Venkat Poornalingam写道: