| 1 |
|
|---|
| 2 | wcsa-ef-ingest.process-ef-json-mode = per-volume
|
|---|
| 3 | #wcsa-ef-ingest.process-ef-json-mode = per-page
|
|---|
| 4 | wcsa-ef-ingest.use-whitelist = true
|
|---|
| 5 | wcsa-ef-ingest.whitelist-filename = file:/hdfsd05/dbbridge/whitelist-peter1.txt
|
|---|
| 6 | #wcsa-ef-ingest.whitelist-filename = file:///home/dbbridge/extracted-features-solr/solr-ingest/whitelist-peter1.txt
|
|---|
| 7 |
|
|---|
| 8 |
|
|---|
| 9 | #wcsa-ef-ingest.solr-clode-nodes = 10.11.0.53:8983,10.11.0.54:8983,10.11.0.55:8983
|
|---|
| 10 | wcsa-ef-ingest.solr-cloud-nodes = gc0:8983,gc1:8983,gc2:8983,gc3:8983,gc4:8983,gc5:8983,gc6:8983,gc7:8983,gc8:8983,gc9:8983
|
|---|
| 11 | wcsa-ef-ingest.icu-tokenize = true
|
|---|
| 12 | wcsa-ef-ingest.strict-file-io = false
|
|---|
| 13 |
|
|---|
| 14 |
|
|---|
| 15 | # For guide on number of partitions to use, see "Parallelized collections" section of:
|
|---|
| 16 | # https://spark.apache.org/docs/2.0.1/programming-guide.html
|
|---|
| 17 | # which suggests 2-4 * num_cores
|
|---|
| 18 | #
|
|---|
| 19 | # For a more detailed discussion see:
|
|---|
| 20 | # http://blog.cloudera.com/blog/2015/03/how-to-tune-your-apache-spark-jobs-part-2/
|
|---|
| 21 |
|
|---|
| 22 | # wcsa-ef-ingest.num-partitions = 12
|
|---|
| 23 | #wcsa-ef-ingest.num-partitions = 110
|
|---|
| 24 | #wcsa-ef-ingest.num-partitions = 220
|
|---|
| 25 | #wcsa-ef-ingest.num-partitions = 400
|
|---|
| 26 | #wcsa-ef-ingest.num-partitions = 1000
|
|---|
| 27 | wcsa-ef-ingest.files-per-partition = 1300
|
|---|
| 28 |
|
|---|
| 29 | spark.executor.cores=10
|
|---|
| 30 |
|
|---|
| 31 | spark.driver.memory=50g
|
|---|
| 32 | spark.executor.memory=70g
|
|---|
| 33 | ##spark.network.timeout=240s
|
|---|
| 34 |
|
|---|
| 35 | #spark.local.dir=/var/tmp
|
|---|
