As stated in your other question, what I have seen, mainly when applications are dealing with sensitive data, is:
Data are encrypted and stored securely locally or in a protected server (fog or cloud, for instance);
The data hashes are stored in the blockchain infrastructure.
This is the common flow. To protect data, I have seen the use of Trusted Execution Environments (TEE), such as Intel SGX or Arm Trustzone. The following works approach the mentioned flow in their solutions:
Conference Paper Decentralized IoT Data Management Using BlockChain and Trust...
Conference Paper Towards Decentralized Accountability and Self-Sovereignty in...
The following work presents an architecture that relies on a TEE to protect data in fog/cloud-based IoT applications (although it does not use blockchain, maybe you get some insight that applies also to your blockchain application): Conference Paper Achieving Data Dissemination with Security using FIWARE and ...
Blockchain technology always has an issues to storing large and confidential data. For that, researchers are offloading their data at cloud or edge/fog computing resources and only hashes of metadata are stored in the blockchain. But again, there are various security issues of storing data at cloud that's why decentralized cloud and sycing various cloud vendors for blockchain are addressed. Some researchers are also separate the data storation and computation part for less memory exhaustion. Various popular blockchain platform uses the oracle database for storing the information in nosql form.