Technology Updates
1. Redis integration: The Redis cache layer is introduced to cache hot data, such as frequently queried transactions and block information, reducing direct pressure on the database.
2. Cache invalidation strategy: The LRU algorithm is used to automatically eliminate old cache data to ensure the timeliness of cache layer data.
3. The data consistency between the database and the cache layer is achieved through the Pub/Sub function of Redis. When the data is updated, the relevant cache will automatically expire.
4. The task queue is introduced to synchronize tasks such as data synchronization and smart contract event processing, reducing the blocking of the main thread.
5. The introduction of GraphQL allows the front end to obtain multiple resources with a single request, reducing the number of requests and server load.
6. A composite index based on block_number and transaction_index is created on the transactions table, significantly improving the speed of querying transactions by block.
