Blockchain Developers Face Unexpected Costs with Google Cloud's BigQuery

According to Cointelegraph, blockchain developers are expressing concerns over unexpected high charges from Google Cloud's BigQuery service. A notable incident involved a developer who received a $15,000 bill for conducting three queries. BigQuery, a serverless data warehouse by Google Cloud, is designed for analyzing large datasets using SQL and includes AI capabilities. A pseudonymous developer, whose experience was shared by Mikko Ohtamaa, co-founder of the decentralized algorithmic trading protocol Trading Strategy, described BigQuery as a potential financial risk. The developer typically incurs a few hundred dollars in monthly charges but was shocked to receive an $18,000 bill. The charges stemmed from three queries on Solana, each costing over $5,000. After raising the issue with Google support, the charges were reduced to $4,000 per query. Other members of the crypto community have echoed these concerns, criticizing what they perceive as predatory pricing practices that prevent setting monthly spending limits. Ermin Nurovic, co-founder of the Flat Money synthetic dollar protocol, highlighted the lack of hard stops, noting that a stuck Google Cloud function could lead to exorbitant costs. The integration of Solana with Google Cloud’s BigQuery in October 2023 was intended to facilitate access to Solana blockchain data, such as whale transactions and NFT sales. However, the billing issues have raised concerns among developers. Another developer reported a $5,000 charge for a single query that inadvertently scanned multiple terabytes of data. Fortunately, their company’s local connection to Google allowed them to escalate the issue and secure a refund. These incidents have led developers to exercise caution when querying blockchain data on BigQuery, emphasizing the need to check data partitions beforehand. The pricing model has also raised doubts about the feasibility of using BigQuery for AI algorithms, given the potential for unexpected costs.