ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

[PPML] TPC-DS doc upadte (#6238)

Browse source 
* ppml tpcds doc update

* fix

* update data generation step
This commit is contained in:
CharleneHu94 2022-10-21 17:01:37 +08:00 committed by GitHub
parent f7e07ecc69
commit 5e4c269a49
1 changed files with 189 additions and 182 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

371
docs/readthedocs/source/doc/PPML/QuickStart/tpc-ds_with_sparksql_on_k8s.md
View file

@ -8,207 +8,214 @@
### Prepare TPC-DS kit and data ### Prepare TPC-DS kit and data
1. Download and compile tpc-ds 1. Download and compile TPC-DS kit
```bash ```bash
git clone --recursive https://github.com/intel-analytics/zoo-tutorials.git git clone --recursive https://github.com/intel-analytics/zoo-tutorials.git
cd /path/to/zoo-tutorials cd zoo-tutorials/tpcds-spark
git clone https://github.com/databricks/tpcds-kit.git git clone https://github.com/databricks/tpcds-kit.git
cd tpcds-kit/tools cd tpcds-kit/tools
make OS=LINUX make OS=LINUX
``` cd ../../
sbt package
```
2. Generate data 2. Generate data
```bash ```bash
cd /path/to/zoo-tutorials cd /path/to/zoo-tutorials/tpcds-spark/spark-sql-perf
cd tpcds-spark/spark-sql-perf sbt "test:runMain com.databricks.spark.sql.perf.tpcds.GenTPCDSData -d <dsdgenDir> -s <scaleFactor> -l <dataDir> -f parquet"
sbt "test:runMain com.databricks.spark.sql.perf.tpcds.GenTPCDSData -d <dsdgenDir> -s <scaleFactor> -l <dataDir> -f parquet" ```
```
`dsdgenDir` is the path of `tpcds-kit/tools`, `scaleFactor` is the size of the data, for example `-s 1` will generate 1G data, `dataDir` is the path to store generated data. `dsdgenDir` is the path of `tpcds-kit/tools`, `scaleFactor` indicates data size, for example `-s 1` will generate data of 1GB scale factor, `dataDir` is the path to store generated data.
### Deploy PPML TPC-DS on Kubernetes ### Deploy PPML TPC-DS on Kubernetes
1. Pull docker image
1. Compile Kit ```bash
sudo docker pull intelanalytics/bigdl-ppml-trusted-big-data-ml-python-graphene:2.1.0-SNAPSHOT
```
```bash 2. Prepare keys, password and k8s configurations (follow instructions [here](https://github.com/intel-analytics/BigDL/tree/main/ppml/trusted-big-data-ml/python/docker-graphene#11-prepare-the-keyspassworddataenclave-keypem "here")), make sure keys, `tpcds-spark` and generated tpc-ds data can be accessed on each K8S node, e.g. deploy on distributed storage inclusing NFS and HDFS.
cd zoo-tutorials/tpcds-spark 3. Start a bigdl-ppml enabled Spark K8S client container with configured local IP, key, tpc-ds and kubeconfig path, also configure data path if your data is stored on local FS
sbt package
```
2. Create external tables ```bash
export ENCLAVE_KEY=/YOUR_DIR/keys/enclave-key.pem
export TPCDS_PATH=/YOUR_DIR/zoo-tutorials/tpcds-spark
export DATA_PATH=/YOUR_DIR/data
export KEYS_PATH=/YOUR_DIR/keys
export SECURE_PASSWORD_PATH=/YOUR_DIR/password
export KUBECONFIG_PATH=/YOUR_DIR/kubeconfig
export LOCAL_IP=$local_ip
export DOCKER_IMAGE=intelanalytics/bigdl-ppml-trusted-big-data-ml-python-graphene:2.1.0-SNAPSHOT
sudo docker run -itd \
--privileged \
--net=host \
--name=spark-k8s-client \
--oom-kill-disable \
--device=/dev/sgx/enclave \
--device=/dev/sgx/provision \
-v /var/run/aesmd/aesm.socket:/var/run/aesmd/aesm.socket \
-v $ENCLAVE_KEY:/graphene/Pal/src/host/Linux-SGX/signer/enclave-key.pem \
-v $TPCDS_PATH:/ppml/trusted-big-data-ml/work/tpcds-spark \
-v $DATA_PATH:/ppml/trusted-big-data-ml/work/data \
-v $KEYS_PATH:/ppml/trusted-big-data-ml/work/keys \
-v $SECURE_PASSWORD_PATH:/ppml/trusted-big-data-ml/work/password \
-v $KUBECONFIG_PATH:/root/.kube/config \
-e RUNTIME_SPARK_MASTER=k8s://https://$LOCAL_IP:6443 \
-e RUNTIME_K8S_SERVICE_ACCOUNT=spark \
-e RUNTIME_K8S_SPARK_IMAGE=$DOCKER_IMAGE \
-e RUNTIME_DRIVER_HOST=$LOCAL_IP \
-e RUNTIME_DRIVER_PORT=54321 \
-e RUNTIME_EXECUTOR_INSTANCES=1 \
-e RUNTIME_EXECUTOR_CORES=4 \
-e RUNTIME_EXECUTOR_MEMORY=20g \
-e RUNTIME_TOTAL_EXECUTOR_CORES=4 \
-e RUNTIME_DRIVER_CORES=4 \
-e RUNTIME_DRIVER_MEMORY=10g \
-e SGX_MEM_SIZE=64G \
-e SGX_LOG_LEVEL=error \
-e LOCAL_IP=$LOCAL_IP \
$DOCKER_IMAGE bash
```
```bash 4. Attach to the client container
$SPARK_HOME/bin/spark-submit \
--class "createTables" \
--master <spark-master> \
--driver-memory 20G \
--executor-cores <executor-cores> \
--total-executor-cores <total-cores> \
--executor-memory 20G \
--jars spark-sql-perf/target/scala-2.12/spark-sql-perf_2.12-0.5.1-SNAPSHOT.jar \
target/scala-2.12/tpcds-benchmark_2.12-0.1.jar <dataDir> <dsdgenDir> <scaleFactor>
```
3. Pull docker image ```bash
sudo docker exec -it spark-local-k8s-client bash
```
```bash 5. Create external tables
sudo docker pull intelanalytics/bigdl-ppml-trusted-big-data-ml-python-graphene:2.1.0-SNAPSHOT
```
4. Prepare SGX keys (following instructions [here](https://github.com/intel-analytics/BigDL/tree/main/ppml/trusted-big-data-ml/python/docker-graphene#11-prepare-the-keyspassworddataenclave-keypem "here")), make sure keys and tpcds-spark can be accessed on each K8S node ```bash
5. Start a bigdl-ppml enabled Spark K8S client container with configured local IP, key, tpc-ds and kuberconfig path cd /ppml/trusted-big-data-ml/work/tpcds-spark
$SPARK_HOME/bin/spark-submit \
--class "createTables" \
--master <spark-master> \
--driver-memory 20G \
--executor-cores <executor-cores> \
--total-executor-cores <total-cores> \
--executor-memory 20G \
--jars spark-sql-perf/target/scala-2.12/spark-sql-perf_2.12-0.5.1-SNAPSHOT.jar \
target/scala-2.12/tpcds-benchmark_2.12-0.1.jar <dataDir> <dsdgenDir> <scaleFactor>
```
`<dataDir>` and `<dsdgenDir>` are the generated data path and `tpcds-kit/tools` path, both should be accessible in the container. After successfully creating tables, there should be a directory `metastore_db` in the current working path.
```bash 6. Modify `/ppml/trusted-big-data-ml/spark-executor-template.yaml`, add path of `enclave-key`, `tpcds-spark` and `kubeconfig`. If data is not stored on HDFS, also configure mount volume `data` and make sure `mountPath` is the same as `<dataDir>` used in create table step.
export ENCLAVE_KEY=/YOUR_DIR/keys/enclave-key.pem
export DATA_PATH=/YOUR_DIR/zoo-tutorials/tpcds-spark
export KEYS_PATH=/YOUR_DIR/keys
export SECURE_PASSWORD_PATH=/YOUR_DIR/password
export KUBERCONFIG_PATH=/YOUR_DIR/kuberconfig
export LOCAL_IP=$local_ip
export DOCKER_IMAGE=intelanalytics/bigdl-ppml-trusted-big-data-ml-python-graphene:2.1.0-SNAPSHOT
sudo docker run -itd \
--privileged \
--net=host \
--name=spark-local-k8s-client \
--oom-kill-disable \
--device=/dev/sgx/enclave \
--device=/dev/sgx/provision \
-v /var/run/aesmd/aesm.socket:/var/run/aesmd/aesm.socket \
-v $ENCLAVE_KEY:/graphene/Pal/src/host/Linux-SGX/signer/enclave-key.pem \
-v $DATA_PATH:/ppml/trusted-big-data-ml/work/tpcds-spark \
-v $KEYS_PATH:/ppml/trusted-big-data-ml/work/keys \
-v $SECURE_PASSWORD_PATH:/ppml/trusted-big-data-ml/work/password \
-v $KUBERCONFIG_PATH:/root/.kube/config \
-e RUNTIME_SPARK_MASTER=k8s://https://$LOCAL_IP:6443 \
-e RUNTIME_K8S_SERVICE_ACCOUNT=spark \
-e RUNTIME_K8S_SPARK_IMAGE=$DOCKER_IMAGE \
-e RUNTIME_DRIVER_HOST=$LOCAL_IP \
-e RUNTIME_DRIVER_PORT=54321 \
-e RUNTIME_EXECUTOR_INSTANCES=1 \
-e RUNTIME_EXECUTOR_CORES=4 \
-e RUNTIME_EXECUTOR_MEMORY=20g \
-e RUNTIME_TOTAL_EXECUTOR_CORES=4 \
-e RUNTIME_DRIVER_CORES=4 \
-e RUNTIME_DRIVER_MEMORY=10g \
-e SGX_MEM_SIZE=64G \
-e SGX_LOG_LEVEL=error \
-e LOCAL_IP=$LOCAL_IP \
$DOCKER_IMAGE bash
```
6. Attach to the client container ```yaml
apiVersion: v1
```bash kind: Pod
sudo docker exec -it spark-local-k8s-client bash spec:
``` containers:
- name: spark-executor
7. Modify `spark-executor-template.yaml`, add path of `enclave-key`, `tpcds-spark` and `kuberconfig` on host securityContext:
privileged: true
```yaml volumeMounts:
apiVersion: v1 - name: enclave-key
kind: Pod mountPath: /graphene/Pal/src/host/Linux-SGX/signer/enclave-key.pem
spec:
containers:
- name: spark-executor
securityContext:
privileged: true
volumeMounts:
... ...
- name: tpcds - name: tpcds
mountPath: /ppml/trusted-big-data-ml/work/tpcds-spark mountPath: /ppml/trusted-big-data-ml/work/tpcds-spark
- name: kubeconf - name: data
mountPath: /root/.kube/config mountPath: /mounted/path/to/data
volumes: - name: kubeconf
- name: enclave-key mountPath: /root/.kube/config
hostPath: volumes:
path: /root/keys/enclave-key.pem - name: enclave-key
hostPath:
path: /path/to/keys/enclave-key.pem
... ...
- name: tpcds - name: tpcds
hostPath: hostPath:
path: /path/to/tpcds-spark path: /path/to/tpcds-spark
- name: kubeconf - name: data
hostPath: hostPath:
path: /path/to/kuberconfig path: /path/to/data
``` - name: kubeconf
hostPath:
path: /path/to/kubeconfig
```
8. Execute TPC-DS queries 7. Execute TPC-DS queries
Optional argument `QUERY` is the query number to run. Multiple query numbers should be separated by space, e.g. `1 2 3`. If no query number is specified, all 1-99 queries would be executed. Optional argument `QUERY` is the query number to run. Multiple query numbers should be separated by space, e.g. `1 2 3`. If no query number is specified, all 1-99 queries would be executed. Configure `$hdfs_host_ip` and `$hdfs_port` if the output is stored on HDFS.
```bash ```bash
secure_password=`openssl rsautl -inkey /ppml/trusted-big-data-ml/work/password/key.txt -decrypt </ppml/trusted-big-data-ml/work/password/output.bin` && \ cd /ppml/trusted-big-data-ml/work/tpcds-spark
export TF_MKL_ALLOC_MAX_BYTES=10737418240 && \ secure_password=`openssl rsautl -inkey /ppml/trusted-big-data-ml/work/password/key.txt -decrypt </ppml/trusted-big-data-ml/work/password/output.bin` && \
export SPARK_LOCAL_IP=$LOCAL_IP && \ export TF_MKL_ALLOC_MAX_BYTES=10737418240 && \
export HDFS_HOST=$hdfs_host_ip && \ export SPARK_LOCAL_IP=$LOCAL_IP && \
export HDFS_PORT=$hdfs_port && \ export HDFS_HOST=$hdfs_host_ip && \
export TPCDS_DIR=/ppml/trusted-big-data-ml/work/tpcds-spark \ export HDFS_PORT=$hdfs_port && \
export OUTPUT_DIR=hdfs://$HDFS_HOST:$HDFS_PORT/tpc-ds/output \ export TPCDS_DIR=/ppml/trusted-big-data-ml/work/tpcds-spark \
export QUERY=3 export OUTPUT_DIR=hdfs://$HDFS_HOST:$HDFS_PORT/tpc-ds/output \
/opt/jdk8/bin/java \ export QUERY=3
-cp '$TPCDS_DIR/target/scala-2.12/tpcds-benchmark_2.12-0.1.jar:/ppml/trusted-big-data-ml/work/spark-3.1.2/conf/:/ppml/trusted-big-data-ml/work/spark-3.1.2/jars/*' \ /opt/jdk8/bin/java \
-Xmx10g \ -cp '$TPCDS_DIR/target/scala-2.12/tpcds-benchmark_2.12-0.1.jar:$TPCDS_DIR/spark-sql-perf/target/scala-2.12/spark-sql-perf_2.12-0.5.1-SNAPSHOT.jar:/ppml/trusted-big-data-ml/work/spark-3.1.2/conf/:/ppml/trusted-big-data-ml/work/spark-3.1.2/jars/*' \
-Dbigdl.mklNumThreads=1 \ -Xmx10g \
org.apache.spark.deploy.SparkSubmit \ -Dbigdl.mklNumThreads=1 \
--master $RUNTIME_SPARK_MASTER \ org.apache.spark.deploy.SparkSubmit \
--deploy-mode client \ --master $RUNTIME_SPARK_MASTER \
--name spark-tpcds-sgx \ --deploy-mode client \
--conf spark.driver.host=$LOCAL_IP \ --name spark-tpcds-sgx \
--conf spark.driver.port=54321 \ --conf spark.driver.host=$LOCAL_IP \
--conf spark.driver.memory=10g \ --conf spark.driver.port=54321 \
--conf spark.driver.blockManager.port=10026 \ --conf spark.driver.memory=10g \
--conf spark.blockManager.port=10025 \ --conf spark.driver.blockManager.port=10026 \
--conf spark.scheduler.maxRegisteredResourcesWaitingTime=5000000 \ --conf spark.blockManager.port=10025 \
--conf spark.worker.timeout=600 \ --conf spark.scheduler.maxRegisteredResourcesWaitingTime=5000000 \
--conf spark.python.use.daemon=false \ --conf spark.worker.timeout=600 \
--conf spark.python.worker.reuse=false \ --conf spark.python.use.daemon=false \
--conf spark.network.timeout=10000000 \ --conf spark.python.worker.reuse=false \
--conf spark.starvation.timeout=250000 \ --conf spark.network.timeout=10000000 \
--conf spark.rpc.askTimeout=600 \ --conf spark.starvation.timeout=250000 \
--conf spark.sql.autoBroadcastJoinThreshold=-1 \ --conf spark.rpc.askTimeout=600 \
--conf spark.io.compression.codec=lz4 \ --conf spark.sql.autoBroadcastJoinThreshold=-1 \
--conf spark.sql.shuffle.partitions=8 \ --conf spark.io.compression.codec=lz4 \
--conf spark.speculation=false \ --conf spark.sql.shuffle.partitions=8 \
--conf spark.executor.heartbeatInterval=10000000 \ --conf spark.speculation=false \
--conf spark.executor.instances=24 \ --conf spark.executor.heartbeatInterval=10000000 \
--executor-cores 8 \ --conf spark.executor.instances=24 \
--total-executor-cores 192 \ --executor-cores 8 \
--executor-memory 16G \ --total-executor-cores 192 \
--properties-file /ppml/trusted-big-data-ml/work/bigdl-2.1.0-SNAPSHOT/conf/spark-bigdl.conf \ --executor-memory 16G \
--conf spark.kubernetes.authenticate.serviceAccountName=spark \ --properties-file /ppml/trusted-big-data-ml/work/bigdl-2.1.0-SNAPSHOT/conf/spark-bigdl.conf \
--conf spark.kubernetes.container.image=$RUNTIME_K8S_SPARK_IMAGE \ --conf spark.kubernetes.authenticate.serviceAccountName=spark \
--conf spark.kubernetes.executor.podTemplateFile=/ppml/trusted-big-data-ml/spark-executor-template.yaml \ --conf spark.kubernetes.container.image=$RUNTIME_K8S_SPARK_IMAGE \
--conf spark.kubernetes.executor.deleteOnTermination=false \ --conf spark.kubernetes.executor.podTemplateFile=/ppml/trusted-big-data-ml/spark-executor-template.yaml \
--conf spark.kubernetes.executor.podNamePrefix=spark-tpcds-sgx \ --conf spark.kubernetes.executor.deleteOnTermination=false \
--conf spark.kubernetes.sgx.enabled=true \ --conf spark.kubernetes.executor.podNamePrefix=spark-tpcds-sgx \
--conf spark.kubernetes.sgx.executor.mem=32g \ --conf spark.kubernetes.sgx.enabled=true \
--conf spark.kubernetes.sgx.executor.jvm.mem=6g \ --conf spark.kubernetes.sgx.executor.mem=32g \
--conf spark.kubernetes.sgx.log.level=$SGX_LOG_LEVEL \ --conf spark.kubernetes.sgx.executor.jvm.mem=6g \
--conf spark.authenticate=true \ --conf spark.kubernetes.sgx.log.level=$SGX_LOG_LEVEL \
--conf spark.authenticate.secret=$secure_password \ --conf spark.authenticate=true \
--conf spark.kubernetes.executor.secretKeyRef.SPARK_AUTHENTICATE_SECRET="spark-secret:secret" \ --conf spark.authenticate.secret=$secure_password \
--conf spark.kubernetes.driver.secretKeyRef.SPARK_AUTHENTICATE_SECRET="spark-secret:secret" \ --conf spark.kubernetes.executor.secretKeyRef.SPARK_AUTHENTICATE_SECRET="spark-secret:secret" \
--conf spark.authenticate.enableSaslEncryption=true \ --conf spark.kubernetes.driver.secretKeyRef.SPARK_AUTHENTICATE_SECRET="spark-secret:secret" \
--conf spark.network.crypto.enabled=true \ --conf spark.authenticate.enableSaslEncryption=true \
--conf spark.network.crypto.keyLength=128 \ --conf spark.network.crypto.enabled=true \
--conf spark.network.crypto.keyFactoryAlgorithm=PBKDF2WithHmacSHA1 \ --conf spark.network.crypto.keyLength=128 \
--conf spark.io.encryption.enabled=true \ --conf spark.network.crypto.keyFactoryAlgorithm=PBKDF2WithHmacSHA1 \
--conf spark.io.encryption.keySizeBits=128 \ --conf spark.io.encryption.enabled=true \
--conf spark.io.encryption.keygen.algorithm=HmacSHA1 \ --conf spark.io.encryption.keySizeBits=128 \
--conf spark.ssl.enabled=true \ --conf spark.io.encryption.keygen.algorithm=HmacSHA1 \
--conf spark.ssl.port=8043 \ --conf spark.ssl.enabled=true \
--conf spark.ssl.keyPassword=$secure_password \ --conf spark.ssl.port=8043 \
--conf spark.ssl.keyStore=/ppml/trusted-big-data-ml/work/keys/keystore.jks \ --conf spark.ssl.keyPassword=$secure_password \
--conf spark.ssl.keyStorePassword=$secure_password \ --conf spark.ssl.keyStore=/ppml/trusted-big-data-ml/work/keys/keystore.jks \
--conf spark.ssl.keyStoreType=JKS \ --conf spark.ssl.keyStorePassword=$secure_password \
--conf spark.ssl.trustStore=/ppml/trusted-big-data-ml/work/keys/keystore.jks \ --conf spark.ssl.keyStoreType=JKS \
--conf spark.ssl.trustStorePassword=$secure_password \ --conf spark.ssl.trustStore=/ppml/trusted-big-data-ml/work/keys/keystore.jks \
--conf spark.ssl.trustStoreType=JKS \ --conf spark.ssl.trustStorePassword=$secure_password \
--class "TPCDSBenchmark" \ --conf spark.ssl.trustStoreType=JKS \
--verbose \ --class "TPCDSBenchmark" \
$TPCDS_DIR/target/scala-2.12/tpcds-benchmark_2.12-0.1.jar \ --jars $TPCDS_DIR/spark-sql-perf/target/scala-2.12/spark-sql-perf_2.12-0.5.1-SNAPSHOT.jar \
$OUTPUT_DIR $QUERY --verbose \
``` $TPCDS_DIR/target/scala-2.12/tpcds-benchmark_2.12-0.1.jar \
$OUTPUT_DIR $QUERY
```
Note: For Spark cluster mode, the `metastore_db` directory generated in table create step needs to be mounted into driver pod, and the path in the container needs to specified by adding `--conf spark.hadoop.javax.jdo.option.ConnectionURL="jdbc:derby:;databaseName=/path/to/metastore_db;create=true" \` to `spark-submit` command.
After benchmark is finished, the performance result is saved as `part-*.csv` file under `<OUTPUT_DIR>/performance` directory. After benchmark is finished, the performance result is saved as `part-*.csv` file under `<OUTPUT_DIR>/performance` directory.