summaryrefslogtreecommitdiff
path: root/scripts/05-dataproc-submit.sh
diff options
context:
space:
mode:
authorSanto Cariotti <santo@dcariotti.me>2024-12-27 22:22:35 +0100
committerSanto Cariotti <santo@dcariotti.me>2024-12-27 22:22:35 +0100
commit299f5ab9c38834fc58b2f2a434c1495ac3d1c554 (patch)
tree319dbf5aa7a9d507b985760e860934484002ca4b /scripts/05-dataproc-submit.sh
parentfc35b02c84c94d5965a8317df6312c9a3ececc68 (diff)
Add scripts
Diffstat (limited to 'scripts/05-dataproc-submit.sh')
-rwxr-xr-xscripts/05-dataproc-submit.sh9
1 files changed, 9 insertions, 0 deletions
diff --git a/scripts/05-dataproc-submit.sh b/scripts/05-dataproc-submit.sh
new file mode 100755
index 0000000..dfc5498
--- /dev/null
+++ b/scripts/05-dataproc-submit.sh
@@ -0,0 +1,9 @@
+#!/bin/sh
+
+gcloud dataproc jobs submit spark \
+ --cluster=${CLUSTER} \
+ --jar=gs://${BUCKET_NAME}/scala/co-purchase-analysis_2.12-1.0.jar \
+ --region=${REGION} \
+ --properties spark.hadoop.fs.gs.impl=com.google.cloud.hadoop.fs.gcs.GoogleHadoopFileSystem \
+ -- gs://${BUCKET_NAME}/input/ gs://${BUCKET_NAME}/output/
+