Bỏ qua đến nội dung chính
Back to home
Tech AI 2 min read

Vercel Blob now supports consistent reads on private storage

Vercel Blob has launched consistent reads for private storage, allowing developers to fetch the latest data instantly by bypassing the CDN cache.

Tier 2 · sources 50% confidence Reviewed
Sources vercel.com

Vercel has officially added consistent reads support for Vercel Blob on private storage partitions. This new update addresses the real-time data synchronization challenge by allowing developers to bypass the CDN cache when necessary. This is an important technical improvement for applications requiring real-time data accuracy.

Bối cảnh & Nguyên nhân

Normally, when a blob is overwritten on an existing pathname, subsequent read requests might still receive the cached version from the Content Delivery Network (CDN) for up to 60 seconds. For static files like images or general documents, this delay is perfectly acceptable. However, for dynamic applications requiring continuous updates such as AI agent memories, chat transcripts, or scheduled JSON reports, reading outdated data for up to a minute can cause severe logic errors in the system.

Phân tích kỹ thuật & Công nghệ

To resolve this issue, Vercel provides a direct intervention mechanism in the data query process. In the latest @vercel/blob SDK, developers simply need to pass useCache: false to the get() or presignUrl() functions. Under the hood, setting useCache: false automatically appends a cache=0 query parameter to the private blob URL to bypass the CDN and fetch directly from the origin server. In return, these non-cached reads will take longer to respond and will incur Fast Origin Transfer charges.

Ý kiến chuyên gia & Nhận định

According to technical documentation from Vercel, this feature is specifically designed for time-sensitive tasks. Web development experts note that allowing the flexibility to bypass caching gives developers better control over the trade-off between performance (CDN speed) and data consistency. Users do not necessarily have to use Vercel's SDK and can still manually configure the cache=0 parameter directly on private blob URLs.

Tác động & Tương lai

This improvement to Vercel Blob reflects the strong growth trend of AI agent applications and automated systems that require state memory to be updated and retrieved instantly. For the developer community building SaaS solutions or integrating AI on the Vercel platform, upgrading to the latest SDK will significantly improve application reliability, avoiding common data desynchronization issues experienced in the past.