1
0

[HUDI-2883] Refactor hive sync tool / config to use reflection and standardize configs (#4175)

- Refactor hive sync tool / config to use reflection and standardize configs

Co-authored-by: sivabalan <n.siva.b@gmail.com>
Co-authored-by: Rajesh Mahindra <rmahindra@Rajeshs-MacBook-Pro.local>
Co-authored-by: Raymond Xu <2701446+xushiyan@users.noreply.github.com>
This commit is contained in:
Rajesh Mahindra
2022-03-21 19:56:31 -07:00
committed by GitHub
parent 9b6e138af2
commit 5f570ea151
43 changed files with 1521 additions and 1217 deletions

View File

@@ -21,6 +21,8 @@ import org.apache.hudi.DataSourceWriteOptions;
import org.apache.spark.sql.SaveMode;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.HoodieDataSourceHelpers;
import org.apache.hudi.hive.HiveSyncConfig;
import org.apache.hudi.sync.common.HoodieSyncConfig;
import org.apache.hudi.hive.MultiPartKeysValueExtractor;
import org.apache.hadoop.fs.FileSystem;
@@ -43,14 +45,14 @@ spark.sql("select key, `_hoodie_partition_path` as datestr, symbol, ts, open, cl
option(DataSourceWriteOptions.PARTITIONPATH_FIELD.key(), "datestr").
option(DataSourceWriteOptions.PRECOMBINE_FIELD.key(), "ts").
option(HoodieWriteConfig.TBL_NAME.key(), "stock_ticks_derived_mor").
option(DataSourceWriteOptions.HIVE_TABLE.key(), "stock_ticks_derived_mor").
option(DataSourceWriteOptions.HIVE_DATABASE.key(), "default").
option(DataSourceWriteOptions.HIVE_URL.key(), "jdbc:hive2://hiveserver:10000").
option(DataSourceWriteOptions.HIVE_USER.key(), "hive").
option(DataSourceWriteOptions.HIVE_PASS.key(), "hive").
option(DataSourceWriteOptions.HIVE_SYNC_ENABLED.key(), "true").
option(DataSourceWriteOptions.HIVE_PARTITION_FIELDS.key(), "datestr").
option(DataSourceWriteOptions.HIVE_PARTITION_EXTRACTOR_CLASS.key(), classOf[MultiPartKeysValueExtractor].getCanonicalName).
option(HoodieSyncConfig.META_SYNC_TABLE_NAME.key(), "stock_ticks_derived_mor").
option(HoodieSyncConfig.META_SYNC_DATABASE_NAME.key(), "default").
option(HiveSyncConfig.HIVE_URL.key(), "jdbc:hive2://hiveserver:10000").
option(HiveSyncConfig.HIVE_USER.key(), "hive").
option(HiveSyncConfig.HIVE_PASS.key(), "hive").
option(HiveSyncConfig.HIVE_SYNC_ENABLED.key(), "true").
option(HoodieSyncConfig.META_SYNC_PARTITION_FIELDS.key(), "datestr").
option(HoodieSyncConfig.META_SYNC_PARTITION_EXTRACTOR_CLASS.key(), classOf[MultiPartKeysValueExtractor].getCanonicalName).
option(DataSourceWriteOptions.URL_ENCODE_PARTITIONING.key(), "true").
mode(SaveMode.Overwrite).
save("/user/hive/warehouse/stock_ticks_derived_mor");
@@ -75,14 +77,14 @@ spark.sql("select key, `_hoodie_partition_path` as datestr, symbol, ts, open, cl
option(DataSourceWriteOptions.PARTITIONPATH_FIELD.key(), "datestr").
option(DataSourceWriteOptions.PRECOMBINE_FIELD.key(), "ts").
option(HoodieWriteConfig.TBL_NAME.key(), "stock_ticks_derived_mor_bs").
option(DataSourceWriteOptions.HIVE_TABLE.key(), "stock_ticks_derived_mor_bs").
option(DataSourceWriteOptions.HIVE_DATABASE.key(), "default").
option(DataSourceWriteOptions.HIVE_URL.key(), "jdbc:hive2://hiveserver:10000").
option(DataSourceWriteOptions.HIVE_USER.key(), "hive").
option(DataSourceWriteOptions.HIVE_PASS.key(), "hive").
option(DataSourceWriteOptions.HIVE_SYNC_ENABLED.key(), "true").
option(DataSourceWriteOptions.HIVE_PARTITION_FIELDS.key(), "datestr").
option(DataSourceWriteOptions.HIVE_PARTITION_EXTRACTOR_CLASS.key(), classOf[MultiPartKeysValueExtractor].getCanonicalName).
option(HoodieSyncConfig.META_SYNC_TABLE_NAME.key(), "stock_ticks_derived_mor_bs").
option(HoodieSyncConfig.META_SYNC_DATABASE_NAME.key(), "default").
option(HiveSyncConfig.HIVE_URL.key(), "jdbc:hive2://hiveserver:10000").
option(HiveSyncConfig.HIVE_USER.key(), "hive").
option(HiveSyncConfig.HIVE_PASS.key(), "hive").
option(HiveSyncConfig.HIVE_SYNC_ENABLED.key(), "true").
option(HoodieSyncConfig.META_SYNC_PARTITION_FIELDS.key(), "datestr").
option(HoodieSyncConfig.META_SYNC_PARTITION_EXTRACTOR_CLASS.key(), classOf[MultiPartKeysValueExtractor].getCanonicalName).
option(DataSourceWriteOptions.URL_ENCODE_PARTITIONING.key(), "true").
mode(SaveMode.Overwrite).
save("/user/hive/warehouse/stock_ticks_derived_mor_bs");