I'm not sure it was dataflow-runner issue (quite likely it wasn't), but want to explore it further.
I tried to run simple Spark job with following cluster specification:
{
"schema": "iglu:com.snowplowanalytics.dataflowrunner/ClusterConfig/avro/1-1-0",
"data": {
"name": "dataflow-runner - snowflake transformer",
"logUri": "s3://mylogs/logs/",
"region": "us-east-1",
"credentials": {
"accessKeyId": "env",
"secretAccessKey": "env"
},
"roles": {
"jobflow": "EMR_EC2_DefaultRole",
"service": "EMR_DefaultRole"
},
"ec2": {
"amiVersion": "5.5.0",
"keyName": "mykey",
"location": {
"classic": {
"availabilityZone": "us-east-1a"
}
},
"instances": {
"master": {
"type": "m1.medium"
},
"core": {
"type": "m1.medium",
"count": 1
},
"task": {
"type": "m1.medium",
"count": 0,
"bid": "0.015"
}
}
},
"tags": [ ],
"bootstrapActionConfigs": [ ],
"configurations": [
{
"classification": "core-site",
"properties": {
"Io.file.buffer.size": "65536"
}
},
{
"classification": "mapred-site",
"properties": {
"Mapreduce.user.classpath.first": "true"
}
}
],
"applications": [ "Hadoop", "Spark" ]
}
}
However, no matter what I submitted - job hung in RUNNING state, not producing any output. To make it work I made two changes:
- Changed
location to {"vpc": null}
- Bumped core and master to
m2.xlarge
Not sure what from above worked and whether it was EMR misconfiguration or something from dataflow-runner, but maybe it's worth for raising.
I'm not sure it was dataflow-runner issue (quite likely it wasn't), but want to explore it further.
I tried to run simple Spark job with following cluster specification:
{ "schema": "iglu:com.snowplowanalytics.dataflowrunner/ClusterConfig/avro/1-1-0", "data": { "name": "dataflow-runner - snowflake transformer", "logUri": "s3://mylogs/logs/", "region": "us-east-1", "credentials": { "accessKeyId": "env", "secretAccessKey": "env" }, "roles": { "jobflow": "EMR_EC2_DefaultRole", "service": "EMR_DefaultRole" }, "ec2": { "amiVersion": "5.5.0", "keyName": "mykey", "location": { "classic": { "availabilityZone": "us-east-1a" } }, "instances": { "master": { "type": "m1.medium" }, "core": { "type": "m1.medium", "count": 1 }, "task": { "type": "m1.medium", "count": 0, "bid": "0.015" } } }, "tags": [ ], "bootstrapActionConfigs": [ ], "configurations": [ { "classification": "core-site", "properties": { "Io.file.buffer.size": "65536" } }, { "classification": "mapred-site", "properties": { "Mapreduce.user.classpath.first": "true" } } ], "applications": [ "Hadoop", "Spark" ] } }However, no matter what I submitted - job hung in
RUNNINGstate, not producing any output. To make it work I made two changes:locationto{"vpc": null}m2.xlargeNot sure what from above worked and whether it was EMR misconfiguration or something from dataflow-runner, but maybe it's worth for raising.