Ethereum ETL Airflow

Read this article: https://cloud.google.com/blog/products/data-analytics/ethereum-bigquery-how-we-built-dataset

Setting up Airflow DAGs using Google Cloud Composer

Create BigQuery Datasets

• Sign in to BigQuery https://bigquery.cloud.google.com/
• Create new datasets called crypto_ethereum, crypto_ethereum_raw, crypto_ethereum_temp

Create Google Cloud Storage bucket

• Create a new Google Storage bucket to store exported files https://console.cloud.google.com/storage/browser

Create Google Cloud Composer environment

Create environment here, https://console.cloud.google.com/composer, use Python version 3, In PYPI Packages tab add ethereum-etl==1.2.3.

Create variables in Airflow (Admin > Variables in the UI):

Variable	Description
ethereum_output_bucket	GCS bucket to store exported files
ethereum_provider_uri	URI of Ethereum node
ethereum_destination_dataset_project_id	Project ID of BigQuery datasets
notification_emails	email for notifications

Check other variables in dags/variables.py.

Upload DAGs

> ./upload_dags.sh <airflow_bucket>

Miscellaneous

To upload CSVs to BigQuery:

• Install Google Cloud SDK https://cloud.google.com/sdk/docs/quickstart-debian-ubuntu

• Create a new Google Storage bucket https://console.cloud.google.com/storage/browser

• Upload the files:

> cd output
> gsutil -m rsync -r . gs://<your_bucket>/ethereumetl/export

• Sign in to BigQuery https://bigquery.cloud.google.com/

• Create a new dataset called ethereum_blockchain_raw and ethereum_blockchain

• Load the files from the bucket to BigQuery:

> git clone https://github.com/medvedev1088/ethereum-etl-airflow.git
> cd ethereum-etl-airflow/dags/resources/stages
> bq --location=US load --replace --source_format=CSV --skip_leading_rows=1 ethereum_blockchain_raw.blocks gs://<your_bucket>/ethereumetl/export/blocks/*.csv ./raw/schemas/blocks.json
> bq --location=US load --replace --source_format=CSV --skip_leading_rows=1 ethereum_blockchain_raw.transactions gs://<your_bucket>/ethereumetl/export/transactions/*.csv ./raw/schemas/transactions.json
> bq --location=US load --replace --source_format=CSV --skip_leading_rows=1 ethereum_blockchain_raw.token_transfers gs://<your_bucket>/ethereumetl/export/token_transfers/*.csv ./raw/schemas/token_transfers.json
> bq --location=US load --replace --source_format=CSV --skip_leading_rows=1 ethereum_blockchain_raw.receipts gs://<your_bucket>/ethereumetl/export/receipts/*.csv ./raw/schemas/receipts.json
> bq --location=US load --replace --source_format=NEWLINE_DELIMITED_JSON ethereum_blockchain_raw.logs gs://<your_bucket>/ethereumetl/export/logs/*.json ./raw/schemas/logs.json
> bq --location=US load --replace --source_format=NEWLINE_DELIMITED_JSON ethereum_blockchain_raw.contracts gs://<your_bucket>/ethereumetl/export/contracts/*.json ./raw/schemas/contracts.json
> bq --location=US load --replace --source_format=CSV --skip_leading_rows=1 --allow_quoted_newlines ethereum_blockchain_raw.tokens gs://<your_bucket>/ethereumetl/export/tokens/*.csv ./raw/schemas/tokens.json

Note that NEWLINE_DELIMITED_JSON is used to support REPEATED mode for the columns with lists.

Enrich blocks:

> bq mk --table --description "$(cat ./enrich/descriptions/blocks.txt | tr '\n' ' ')" --time_partitioning_field timestamp ethereum_blockchain.blocks ./enrich/schemas/blocks.json
> bq --location=US query --destination_table ethereum_blockchain.blocks --use_legacy_sql=false "$(cat ./enrich/sqls/blocks.sql | tr '\n' ' ')"

Enrich transactions:

> bq mk --table --description "$(cat ./enrich/descriptions/transactions.txt | tr '\n' ' ')" --time_partitioning_field block_timestamp ethereum_blockchain.transactions ./enrich/schemas/transactions.json
> bq --location=US query --destination_table ethereum_blockchain.transactions --use_legacy_sql=false "$(cat ./enrich/sqls/transactions.sql | tr '\n' ' ')"

Enrich token_transfers:

> bq mk --table --description "$(cat ./enrich/descriptions/token_transfers.txt | tr '\n' ' ')" --time_partitioning_field block_timestamp ethereum_blockchain.token_transfers ./enrich/schemas/token_transfers.json
> bq --location=US query --destination_table ethereum_blockchain.token_transfers --use_legacy_sql=false "$(cat ./enrich/sqls/token_transfers.sql | tr '\n' ' ')"

Enrich logs:

> bq mk --table --description "$(cat ./enrich/descriptions/logs.txt | tr '\n' ' ')" --time_partitioning_field block_timestamp ethereum_blockchain.logs ./enrich/schemas/logs.json
> bq --location=US query --destination_table ethereum_blockchain.logs --use_legacy_sql=false "$(cat ./enrich/sqls/logs.sql | tr '\n' ' ')"

Enrich contracts:

> bq mk --table --description "$(cat ./enrich/descriptions/contracts.txt | tr '\n' ' ')" --time_partitioning_field block_timestamp ethereum_blockchain.contracts ./enrich/schemas/contracts.json
> bq --location=US query --destination_table ethereum_blockchain.contracts --use_legacy_sql=false "$(cat ./enrich/sqls/contracts.sql | tr '\n' ' ')"

Enrich tokens:

> bq mk --table --description "$(cat ./enrich/descriptions/tokens.txt | tr '\n' ' ')" ethereum_blockchain.tokens ./enrich/schemas/tokens.json
> bq --location=US query --destination_table ethereum_blockchain.tokens --use_legacy_sql=false "$(cat ./enrich/sqls/tokens.sql | tr '\n' ' ')"