Browse Source

Modified the data load procedure.

cartershanklin 12 years ago
parent
commit
6f012a11b6

+ 0 - 60
ddl/bin_partitioned/catalog_returns.sql

@@ -39,65 +39,5 @@ create table catalog_returns
     cr_net_loss               float
 )
 partitioned by (cr_returned_date string)
-clustered by (cr_item_sk) sorted by (cr_item_sk) into ${RETURN_BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        cr.cr_returned_date_sk,
-        cr.cr_returned_time_sk,
-        cr.cr_item_sk,
-        cr.cr_refunded_customer_sk,
-        cr.cr_refunded_cdemo_sk,
-        cr.cr_refunded_hdemo_sk,
-        cr.cr_refunded_addr_sk,
-        cr.cr_returning_customer_sk,
-        cr.cr_returning_cdemo_sk,
-        cr.cr_returning_hdemo_sk,
-        cr.cr_returning_addr_sk,
-        cr.cr_call_center_sk,
-        cr.cr_catalog_page_sk,
-        cr.cr_ship_mode_sk,
-        cr.cr_warehouse_sk,
-        cr.cr_reason_sk,
-        cr.cr_order_number,
-        cr.cr_return_quantity,
-        cr.cr_return_amount,
-        cr.cr_return_tax,
-        cr.cr_return_amt_inc_tax,
-        cr.cr_fee,
-        cr.cr_return_ship_cost,
-        cr.cr_refunded_cash,
-        cr.cr_reversed_charge,
-        cr.cr_store_credit,
-        cr.cr_net_loss,
-        dd.d_date as cr_returned_date
-      from ${SOURCE}.catalog_returns cr
-      left outer join ${SOURCE}.date_dim dd
-      on (cr.cr_returned_date_sk = dd.d_date_sk)) tbl
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 1
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 2
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 3
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 4
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 5
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 6
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 7
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 8
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 9
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 10
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 11
-insert overwrite table catalog_returns partition (cr_returned_date) 
-  select * where month(cr_returned_date) = 12
-;

+ 0 - 67
ddl/bin_partitioned/catalog_sales.sql

@@ -46,72 +46,5 @@ create table catalog_sales
     cs_net_profit             float
 )
 partitioned by (cs_sold_date string)
-clustered by (cs_item_sk) sorted by (cs_item_sk) into ${BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        cs.cs_sold_date_sk,
-        cs.cs_sold_time_sk,
-        cs.cs_ship_date_sk,
-        cs.cs_bill_customer_sk,
-        cs.cs_bill_cdemo_sk,
-        cs.cs_bill_hdemo_sk,
-        cs.cs_bill_addr_sk,
-        cs.cs_ship_customer_sk,
-        cs.cs_ship_cdemo_sk,
-        cs.cs_ship_hdemo_sk,
-        cs.cs_ship_addr_sk,
-        cs.cs_call_center_sk,
-        cs.cs_catalog_page_sk,
-        cs.cs_ship_mode_sk,
-        cs.cs_warehouse_sk,
-        cs.cs_item_sk,
-        cs.cs_promo_sk,
-        cs.cs_order_number,
-        cs.cs_quantity,
-        cs.cs_wholesale_cost,
-        cs.cs_list_price,
-        cs.cs_sales_price,
-        cs.cs_ext_discount_amt,
-        cs.cs_ext_sales_price,
-        cs.cs_ext_wholesale_cost,
-        cs.cs_ext_list_price,
-        cs.cs_ext_tax,
-        cs.cs_coupon_amt,
-        cs.cs_ext_ship_cost,
-        cs.cs_net_paid,
-        cs.cs_net_paid_inc_tax,
-        cs.cs_net_paid_inc_ship,
-        cs.cs_net_paid_inc_ship_tax,
-        cs.cs_net_profit,
-        dd.d_date as cs_sold_date
-      from ${SOURCE}.catalog_sales cs
-      left outer join ${SOURCE}.date_dim dd
-      on (cs.cs_sold_date_sk = dd.d_date_sk)) tbl
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 1
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 2
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 3
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 4
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 5
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 6
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 7
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 8
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 9
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 10
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 11
-insert overwrite table catalog_sales partition (cs_sold_date) 
-  select * where month(cs_sold_date) = 12
-;

+ 0 - 32
ddl/bin_partitioned/inventory.sql

@@ -19,37 +19,5 @@ create table inventory
     inv_quantity_on_hand	int
 )
 partitioned by (inv_date string)
-clustered by (inv_item_sk) sorted by (inv_item_sk) into ${BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-  i.inv_date_sk,
-  i.inv_item_sk,
-  i.inv_warehouse_sk,
-  i.inv_quantity_on_hand,
-  d.d_date as inv_date
-  from ${SOURCE}.inventory i
-  left outer join ${SOURCE}.date_dim d
-  on (d.d_date_sk = i.inv_date_sk)) tbl
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '1998' and inv_date < '1998-07'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '1998-07' and inv_date < '1999'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '1999' and inv_date < '1999-07'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '1999-07' and inv_date < '2000'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2000' and inv_date < '2000-07'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2000-07' and inv_date < '2001'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2001' and inv_date < '2001-07'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2001-07' and inv_date < '2002'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2002' and inv_date < '2002-07'
-insert overwrite table inventory partition (inv_date)
-  select * where inv_date >= '2002-07' and inv_date < '2003';
-

+ 0 - 53
ddl/bin_partitioned/store_returns.sql

@@ -32,58 +32,5 @@ create table store_returns
     sr_net_loss               float
 )
 partitioned by (sr_returned_date string)
-clustered by (sr_item_sk) sorted by (sr_item_sk) into ${RETURN_BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        sr.sr_returned_date_sk,
-        sr.sr_return_time_sk,
-        sr.sr_item_sk,
-        sr.sr_customer_sk,
-        sr.sr_cdemo_sk,
-        sr.sr_hdemo_sk,
-        sr.sr_addr_sk,
-        sr.sr_store_sk,
-        sr.sr_reason_sk,
-        sr.sr_ticket_number,
-        sr.sr_return_quantity,
-        sr.sr_return_amt,
-        sr.sr_return_tax,
-        sr.sr_return_amt_inc_tax,
-        sr.sr_fee,
-        sr.sr_return_ship_cost,
-        sr.sr_refunded_cash,
-        sr.sr_reversed_charge,
-        sr.sr_store_credit,
-        sr.sr_net_loss,
-        dd.d_date as sr_returned_date
-      from ${SOURCE}.store_returns sr
-      left outer join ${SOURCE}.date_dim dd
-      on (sr.sr_returned_date_sk = dd.d_date_sk)) tbl
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 1
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 2
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 3
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 4
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 5
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 6
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 7
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 8
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 9
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 10
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 11
-insert overwrite table store_returns partition (sr_returned_date) 
-  select * where month(sr_returned_date) = 12
-;

+ 0 - 56
ddl/bin_partitioned/store_sales.sql

@@ -35,61 +35,5 @@ create table store_sales
     ss_net_profit             float
 )
 partitioned by (ss_sold_date string)
-clustered by (ss_item_sk) sorted by (ss_item_sk) into ${BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        ss.ss_sold_date_sk,
-        ss.ss_sold_time_sk,
-        ss.ss_item_sk,
-        ss.ss_customer_sk,
-        ss.ss_cdemo_sk,
-        ss.ss_hdemo_sk,
-        ss.ss_addr_sk,
-        ss.ss_store_sk,
-        ss.ss_promo_sk,
-        ss.ss_ticket_number,
-        ss.ss_quantity,
-        ss.ss_wholesale_cost,
-        ss.ss_list_price,
-        ss.ss_sales_price,
-        ss.ss_ext_discount_amt,
-        ss.ss_ext_sales_price,
-        ss.ss_ext_wholesale_cost,
-        ss.ss_ext_list_price,
-        ss.ss_ext_tax,
-        ss.ss_coupon_amt,
-        ss.ss_net_paid,
-        ss.ss_net_paid_inc_tax,
-        ss.ss_net_profit,
-        dd.d_date as ss_sold_date
-      from ${SOURCE}.store_sales ss
-      left outer join ${SOURCE}.date_dim dd
-      on (ss.ss_sold_date_sk = dd.d_date_sk)) tbl
-insert overwrite table store_sales partition (ss_sold_date) 
-  select * where month(ss_sold_date) =  1
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 2
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 3
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 4
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 5
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 6
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 7
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 8
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 9
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 10
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 11
-insert overwrite table store_sales partition (ss_sold_date)
-  select * where month(ss_sold_date) = 12
-;

+ 0 - 57
ddl/bin_partitioned/web_returns.sql

@@ -36,62 +36,5 @@ create table web_returns
     wr_net_loss               float
 )
 partitioned by (wr_returned_date string)
-clustered by (wr_item_sk) sorted by (wr_item_sk) into ${RETURN_BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        wr.wr_returned_date_sk,
-        wr.wr_returned_time_sk,
-        wr.wr_item_sk,
-        wr.wr_refunded_customer_sk,
-        wr.wr_refunded_cdemo_sk,
-        wr.wr_refunded_hdemo_sk,
-        wr.wr_refunded_addr_sk,
-        wr.wr_returning_customer_sk,
-        wr.wr_returning_cdemo_sk,
-        wr.wr_returning_hdemo_sk,
-        wr.wr_returning_addr_sk,
-        wr.wr_web_page_sk,
-        wr.wr_reason_sk,
-        wr.wr_order_number,
-        wr.wr_return_quantity,
-        wr.wr_return_amt,
-        wr.wr_return_tax,
-        wr.wr_return_amt_inc_tax,
-        wr.wr_fee,
-        wr.wr_return_ship_cost,
-        wr.wr_refunded_cash,
-        wr.wr_reversed_charge,
-        wr.wr_account_credit,
-        wr.wr_net_loss,
-        dd.d_date as wr_returned_date
-      from ${SOURCE}.web_returns wr
-      left outer join ${SOURCE}.date_dim dd
-      on (wr.wr_returned_date_sk = dd.d_date_sk)) tbl
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 1
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 2
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 3
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 4
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 5
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 6
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 7
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 8
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 9
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 10
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 11
-insert overwrite table web_returns partition (wr_returned_date)
-  select * where month(wr_returned_date) = 12
-;

+ 0 - 67
ddl/bin_partitioned/web_sales.sql

@@ -46,72 +46,5 @@ create table web_sales
     ws_net_profit             float
 )
 partitioned by (ws_sold_date string)
-clustered by (ws_item_sk) sorted by (ws_item_sk) into ${BUCKETS} buckets
 row format serde '${SERDE}'
 stored as ${FILE};
-
-from (select
-        /*+ MAPJOIN(dd) */
-        ws.ws_sold_date_sk,
-        ws.ws_sold_time_sk,
-        ws.ws_ship_date_sk,
-        ws.ws_item_sk,
-        ws.ws_bill_customer_sk,
-        ws.ws_bill_cdemo_sk,
-        ws.ws_bill_hdemo_sk,
-        ws.ws_bill_addr_sk,
-        ws.ws_ship_customer_sk,
-        ws.ws_ship_cdemo_sk,
-        ws.ws_ship_hdemo_sk,
-        ws.ws_ship_addr_sk,
-        ws.ws_web_page_sk,
-        ws.ws_web_site_sk,
-        ws.ws_ship_mode_sk,
-        ws.ws_warehouse_sk,
-        ws.ws_promo_sk,
-        ws.ws_order_number,
-        ws.ws_quantity,
-        ws.ws_wholesale_cost,
-        ws.ws_list_price,
-        ws.ws_sales_price,
-        ws.ws_ext_discount_amt,
-        ws.ws_ext_sales_price,
-        ws.ws_ext_wholesale_cost,
-        ws.ws_ext_list_price,
-        ws.ws_ext_tax,
-        ws.ws_coupon_amt,
-        ws.ws_ext_ship_cost,
-        ws.ws_net_paid,
-        ws.ws_net_paid_inc_tax,
-        ws.ws_net_paid_inc_ship,
-        ws.ws_net_paid_inc_ship_tax,
-        ws.ws_net_profit,
-        dd.d_date as ws_sold_date
-      from ${SOURCE}.web_sales ws
-      left outer join ${SOURCE}.date_dim dd
-      on (ws.ws_sold_date_sk = dd.d_date_sk)) tbl
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 1
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 2
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 3
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 4
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 5
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 6
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 7
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 8
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 9
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 10
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 11
-insert overwrite table web_sales partition (ws_sold_date) 
-  select * where month(ws_sold_date) = 12
-;

+ 31 - 38
tpcds-setup.sh

@@ -16,12 +16,12 @@ if [ $? -ne 0 ]; then
 fi
 
 # Tables in the TPC-DS schema.
-LIST="date_dim time_dim item customer customer_demographics household_demographics customer_address store promotion warehouse ship_mode reason income_band call_center web_page catalog_page inventory store_sales store_returns web_sales web_returns web_site catalog_sales catalog_returns"
+LIST="date_dim time_dim item customer customer_demographics household_demographics customer_address store promotion warehouse ship_mode reason income_band call_center web_page catalog_page web_site"
+FACTS="web_returns store_sales store_returns web_sales catalog_sales catalog_returns inventory"
 
 # Get the parameters.
 SCALE=$1
 DIR=$2
-MODE=$3
 
 # Ensure arguments exist.
 if [ X"$SCALE" = "X" ]; then
@@ -30,9 +30,6 @@ fi
 if [ X"$DIR" = "X" ]; then
 	DIR=/tmp/tpcds-generate
 fi
-if [ X"$MODE" = "X" ]; then
-	MODE=partitioned
-fi
 
 # Sanity checking.
 if [ $SCALE -eq 1 ]; then
@@ -43,9 +40,6 @@ fi
 BUCKETS=13
 RETURN_BUCKETS=1
 SPLIT=16
-STORE_CLAUSES=( "orc" )
-FILE_FORMATS=( "orc" )
-SERDES=( "org.apache.hadoop.hive.ql.io.orc.OrcSerde" )
 
 set -x
 set -e
@@ -55,34 +49,33 @@ hadoop dfs -ls ${DIR}/${SCALE} || (cd tpcds-gen; hadoop jar target/*.jar -d ${DI
 hadoop dfs -ls ${DIR}/${SCALE}
 
 # Generate the text/flat tables. These will be later be converted to ORCFile.
-hive -i settings/load-flat.sql -f ddl/text/alltables.sql -d DB=tpcds_text_${SCALE} -d LOCATION=${DIR}/${SCALE}
+# hive -i settings/load-flat.sql -f ddl/text/alltables.sql -d DB=tpcds_text_${SCALE} -d LOCATION=${DIR}/${SCALE}
 
-# Generate the binary forms of the data.
-if [ $MODE = "partitioned" ]; then
-	i=0
-	for file in "${STORE_CLAUSES[@]}"
-	do
-		for t in ${LIST}
-		do
-			hive -i settings/load-partitioned.sql -f ddl/bin_partitioned/${t}.sql \
-			    -d DB=tpcds_bin_partitioned_${FILE_FORMATS[$i]}_${SCALE} \
-			    -d SOURCE=tpcds_text_${SCALE} -d BUCKETS=${BUCKETS} \
-			    -d RETURN_BUCKETS=${RETURN_BUCKETS} -d FILE="${file}" \
-			    -d SERDE=${SERDES[$i]} -d SPLIT=${SPLIT}
-		done
-		i=$((i+1))
-	done
-else
-	i=0
-	for file in "${STORE_CLAUSES[@]}"
-	do
-		for t in ${LIST}
-		do
-			hive -i settings/load-flat.sql -f ddl/bin_flat/${t}.sql \
-			    -d DB=tpcds_bin_flat_${FILE_FORMATS[$i]}_${SCALE} \
-			    -d SOURCE=tpcds_text_${SCALE} -d FILE="${file}" \
-			    -d SERDE=${SERDES[$i]}
-		done
-	i=$((i+1))
-	done
-fi
+# Populate the smaller tables.
+#for t in ${LIST}
+#do
+#	hive -i settings/load-partitioned.sql -f ddl/bin_partitioned/${t}.sql \
+#	    -d DB=tpcds_bin_partitioned_orc_${SCALE} \
+#	    -d SOURCE=tpcds_text_${SCALE} -d BUCKETS=${BUCKETS} \
+#	    -d RETURN_BUCKETS=${RETURN_BUCKETS} -d FILE="${file}" \
+#	    -d SERDE=org.apache.hadoop.hive.ql.io.orc.OrcSerde -d SPLIT=${SPLIT}
+#done
+
+# Create the partitioned tables.
+for t in ${FACTS}
+do
+	hive -i settings/load-partitioned.sql -f ddl/bin_partitioned/${t}.sql \
+	    -d DB=tpcds_bin_partitioned_orc_${SCALE} \
+	    -d SOURCE=tpcds_text_${SCALE} -d BUCKETS=${BUCKETS} \
+	    -d RETURN_BUCKETS=${RETURN_BUCKETS} -d FILE="${file}" \
+	    -d SERDE=org.apache.hadoop.hive.ql.io.orc.OrcSerde -d SPLIT=${SPLIT}
+done
+
+# Populate the partitioned tables.
+for t in ${FACTS}
+do
+	hadoop jar tpcds-parts-1.0-SNAPSHOT.jar -t ${t}
+	    -i ${DIR}/${t}/
+	    -o /apps/hive/warehouse/tpcds_bin_partitioned_orc_${SCALE}.db/${t}
+	hive -e "msck repair table ${t}"
+done