Prediction.io - pio 训练失败
Prediction.io - pio train fails
我正在使用来自 sphereio/docker-predictionio docker image and the universal recommendation template template-scala-parallel-universal-recommendation 的 Prediction.IO 的 Elasticsearch + Hbase 版本。
pio-start-all
和 pio status
工作正常,事件服务器功能正常。我创建了一个应用程序并导入了数百个事件。
然而,在模板上执行 pio build
之后,pio train
未能给出几个 javax.naming.NameNotFoundException
警告。即使 pio.log
也不包含任何其他内容。
这是我的 engine.json:
{
"comment": " This config file uses default settings for all but the required values see README.md for docs",
"id": "default",
"description": "Default settings",
"engineFactory": "com.test.RecommendationEngine",
"datasource": {
"params": {
"name": "sample-handmade-data.txt",
"appName": "testapp",
"eventNames": ["START"]
}
},
"sparkConf": {
"spark.serializer": "org.apache.spark.serializer.KryoSerializer",
"spark.kryo.registrator": "org.apache.mahout.sparkbindings.io.MahoutKryoRegistrator",
"spark.kryo.referenceTracking": "false",
"spark.kryoserializer.buffer": "300m",
"spark.executor.memory": "4g",
"es.index.auto.create": "true"
},
"algorithms": [{
"comment": "simplest setup where all values are default, popularity based backfill, must add eventsNames",
"name": "ur",
"params": {
"appName": "testapp",
"indexName": "urindex",
"typeName": "items",
"comment": "must have data for the first event or the model will not build, other events are optional",
"eventNames": ["START"]
}
}]
}
并且 pio train
输出:
[INFO] [Console$] Using existing engine manifest JSON at /PredictionIO-0.9.6/engines/universal-recommendation/manifest.json
[INFO] [Runner$] Submission command: /PredictionIO-0.9.6/vendors/spark-1.5.1-bin-hadoop2.6/bin/spark-submit --class io.prediction.workflow.CreateWorkflow --jars file:/PredictionIO-0.9.6/engines/universal-recommendation/target/scala-2.10/template-scala-parallel-universal-recommendation-assembly-0.2.3-deps.jar,file:/PredictionIO-0.9.6/engines/universal-recommendation/target/scala-2.10/template-scala-parallel-universal-recommendation_2.10-0.2.3.jar --files file:/PredictionIO-0.9.6/conf/log4j.properties,file:/PredictionIO-0.9.6/vendors/hbase-1.0.0/conf/hbase-site.xml --driver-class-path /PredictionIO-0.9.6/conf:/PredictionIO-0.9.6/vendors/hbase-1.0.0/conf file:/PredictionIO-0.9.6/lib/pio-assembly-0.9.6.jar --engine-id FYOHZGlAmUH2xAYWNmQFIf9Jls201WVr --engine-version a892fe59be15dcf27a17f07fb76135a967309fda --engine-variant file:/PredictionIO-0.9.6/engines/universal-recommendation/engine.json --verbosity 0 --json-extractor Both --env PIO_STORAGE_SOURCES_HBASE_TYPE=hbase,PIO_ENV_LOADED=1,PIO_STORAGE_REPOSITORIES_METADATA_NAME=pio_meta,PIO_VERSION=0.9.6,PIO_FS_BASEDIR=/root/.pio_store,PIO_STORAGE_SOURCES_ELASTICSEARCH_HOSTS=localhost,PIO_STORAGE_SOURCES_HBASE_HOME=/PredictionIO-0.9.6/vendors/hbase-1.0.0,PIO_HOME=/PredictionIO-0.9.6,PIO_FS_ENGINESDIR=/root/.pio_store/engines,PIO_STORAGE_SOURCES_LOCALFS_PATH=/root/.pio_store/models,PIO_STORAGE_SOURCES_ELASTICSEARCH_TYPE=elasticsearch,PIO_STORAGE_REPOSITORIES_METADATA_SOURCE=ELASTICSEARCH,PIO_STORAGE_REPOSITORIES_MODELDATA_SOURCE=LOCALFS,PIO_STORAGE_REPOSITORIES_EVENTDATA_NAME=pio_event,PIO_STORAGE_SOURCES_ELASTICSEARCH_CLUSTERNAME=predictionio,PIO_STORAGE_SOURCES_ELASTICSEARCH_HOME=/PredictionIO-0.9.6/vendors/elasticsearch-1.4.4,PIO_FS_TMPDIR=/root/.pio_store/tmp,PIO_STORAGE_REPOSITORIES_MODELDATA_NAME=pio_model,PIO_STORAGE_REPOSITORIES_EVENTDATA_SOURCE=HBASE,PIO_CONF_DIR=/PredictionIO-0.9.6/conf,PIO_STORAGE_SOURCES_ELASTICSEARCH_PORTS=9300,PIO_STORAGE_SOURCES_LOCALFS_TYPE=localfs
[INFO] [Engine] Extracting datasource params...
[INFO] [WorkflowUtils$] No 'name' is found. Default empty String will be used.
[INFO] [Engine] Datasource params: (,DataSourceParams(testapp,List(START)))
[INFO] [Engine] Extracting preparator params...
[INFO] [Engine] Preparator params: (,Empty)
[INFO] [Engine] Extracting serving params...
[INFO] [Engine] Serving params: (,Empty)
[INFO] [Remoting] Starting remoting
[INFO] [Remoting] Remoting started; listening on addresses :[akka.tcp://sparkDriver@172.17.0.2:42582]
[WARN] [MetricsSystem] Using default name DAGScheduler for source because spark.app.id is not set.
[INFO] [Engine$] EngineWorkflow.train
[INFO] [Engine$] DataSource: com.test.DataSource@75bd28d
[INFO] [Engine$] Preparator: com.test.Preparator@13278a41
[INFO] [Engine$] AlgorithmList: List(com.test.URAlgorithm@2365ea38)
[INFO] [Engine$] Data sanity check is on.
[WARN] [TableInputFormatBase] Cannot resolve the host name for 9a94fb2890b3/172.17.0.2 because of javax.naming.NameNotFoundException: DNS name not found [response code 3]; remaining name '2.0.17.172.in-addr.arpa'
[INFO] [Engine$] com.test.TrainingData does not support data sanity check. Skipping check.
[WARN] [TableInputFormatBase] Cannot resolve the host name for 9a94fb2890b3/172.17.0.2 because of javax.naming.NameNotFoundException: DNS name not found [response code 3]; remaining name '2.0.17.172.in-addr.arpa'
有一种方法可以解决这个问题。请在启动 docker 容器时使用 google d.n.s。
--dns=8.8.8.8
我正在使用来自 sphereio/docker-predictionio docker image and the universal recommendation template template-scala-parallel-universal-recommendation 的 Prediction.IO 的 Elasticsearch + Hbase 版本。
pio-start-all
和 pio status
工作正常,事件服务器功能正常。我创建了一个应用程序并导入了数百个事件。
然而,在模板上执行 pio build
之后,pio train
未能给出几个 javax.naming.NameNotFoundException
警告。即使 pio.log
也不包含任何其他内容。
这是我的 engine.json:
{
"comment": " This config file uses default settings for all but the required values see README.md for docs",
"id": "default",
"description": "Default settings",
"engineFactory": "com.test.RecommendationEngine",
"datasource": {
"params": {
"name": "sample-handmade-data.txt",
"appName": "testapp",
"eventNames": ["START"]
}
},
"sparkConf": {
"spark.serializer": "org.apache.spark.serializer.KryoSerializer",
"spark.kryo.registrator": "org.apache.mahout.sparkbindings.io.MahoutKryoRegistrator",
"spark.kryo.referenceTracking": "false",
"spark.kryoserializer.buffer": "300m",
"spark.executor.memory": "4g",
"es.index.auto.create": "true"
},
"algorithms": [{
"comment": "simplest setup where all values are default, popularity based backfill, must add eventsNames",
"name": "ur",
"params": {
"appName": "testapp",
"indexName": "urindex",
"typeName": "items",
"comment": "must have data for the first event or the model will not build, other events are optional",
"eventNames": ["START"]
}
}]
}
并且 pio train
输出:
[INFO] [Console$] Using existing engine manifest JSON at /PredictionIO-0.9.6/engines/universal-recommendation/manifest.json
[INFO] [Runner$] Submission command: /PredictionIO-0.9.6/vendors/spark-1.5.1-bin-hadoop2.6/bin/spark-submit --class io.prediction.workflow.CreateWorkflow --jars file:/PredictionIO-0.9.6/engines/universal-recommendation/target/scala-2.10/template-scala-parallel-universal-recommendation-assembly-0.2.3-deps.jar,file:/PredictionIO-0.9.6/engines/universal-recommendation/target/scala-2.10/template-scala-parallel-universal-recommendation_2.10-0.2.3.jar --files file:/PredictionIO-0.9.6/conf/log4j.properties,file:/PredictionIO-0.9.6/vendors/hbase-1.0.0/conf/hbase-site.xml --driver-class-path /PredictionIO-0.9.6/conf:/PredictionIO-0.9.6/vendors/hbase-1.0.0/conf file:/PredictionIO-0.9.6/lib/pio-assembly-0.9.6.jar --engine-id FYOHZGlAmUH2xAYWNmQFIf9Jls201WVr --engine-version a892fe59be15dcf27a17f07fb76135a967309fda --engine-variant file:/PredictionIO-0.9.6/engines/universal-recommendation/engine.json --verbosity 0 --json-extractor Both --env PIO_STORAGE_SOURCES_HBASE_TYPE=hbase,PIO_ENV_LOADED=1,PIO_STORAGE_REPOSITORIES_METADATA_NAME=pio_meta,PIO_VERSION=0.9.6,PIO_FS_BASEDIR=/root/.pio_store,PIO_STORAGE_SOURCES_ELASTICSEARCH_HOSTS=localhost,PIO_STORAGE_SOURCES_HBASE_HOME=/PredictionIO-0.9.6/vendors/hbase-1.0.0,PIO_HOME=/PredictionIO-0.9.6,PIO_FS_ENGINESDIR=/root/.pio_store/engines,PIO_STORAGE_SOURCES_LOCALFS_PATH=/root/.pio_store/models,PIO_STORAGE_SOURCES_ELASTICSEARCH_TYPE=elasticsearch,PIO_STORAGE_REPOSITORIES_METADATA_SOURCE=ELASTICSEARCH,PIO_STORAGE_REPOSITORIES_MODELDATA_SOURCE=LOCALFS,PIO_STORAGE_REPOSITORIES_EVENTDATA_NAME=pio_event,PIO_STORAGE_SOURCES_ELASTICSEARCH_CLUSTERNAME=predictionio,PIO_STORAGE_SOURCES_ELASTICSEARCH_HOME=/PredictionIO-0.9.6/vendors/elasticsearch-1.4.4,PIO_FS_TMPDIR=/root/.pio_store/tmp,PIO_STORAGE_REPOSITORIES_MODELDATA_NAME=pio_model,PIO_STORAGE_REPOSITORIES_EVENTDATA_SOURCE=HBASE,PIO_CONF_DIR=/PredictionIO-0.9.6/conf,PIO_STORAGE_SOURCES_ELASTICSEARCH_PORTS=9300,PIO_STORAGE_SOURCES_LOCALFS_TYPE=localfs
[INFO] [Engine] Extracting datasource params...
[INFO] [WorkflowUtils$] No 'name' is found. Default empty String will be used.
[INFO] [Engine] Datasource params: (,DataSourceParams(testapp,List(START)))
[INFO] [Engine] Extracting preparator params...
[INFO] [Engine] Preparator params: (,Empty)
[INFO] [Engine] Extracting serving params...
[INFO] [Engine] Serving params: (,Empty)
[INFO] [Remoting] Starting remoting
[INFO] [Remoting] Remoting started; listening on addresses :[akka.tcp://sparkDriver@172.17.0.2:42582]
[WARN] [MetricsSystem] Using default name DAGScheduler for source because spark.app.id is not set.
[INFO] [Engine$] EngineWorkflow.train
[INFO] [Engine$] DataSource: com.test.DataSource@75bd28d
[INFO] [Engine$] Preparator: com.test.Preparator@13278a41
[INFO] [Engine$] AlgorithmList: List(com.test.URAlgorithm@2365ea38)
[INFO] [Engine$] Data sanity check is on.
[WARN] [TableInputFormatBase] Cannot resolve the host name for 9a94fb2890b3/172.17.0.2 because of javax.naming.NameNotFoundException: DNS name not found [response code 3]; remaining name '2.0.17.172.in-addr.arpa'
[INFO] [Engine$] com.test.TrainingData does not support data sanity check. Skipping check.
[WARN] [TableInputFormatBase] Cannot resolve the host name for 9a94fb2890b3/172.17.0.2 because of javax.naming.NameNotFoundException: DNS name not found [response code 3]; remaining name '2.0.17.172.in-addr.arpa'
有一种方法可以解决这个问题。请在启动 docker 容器时使用 google d.n.s。
--dns=8.8.8.8