Skip to main content

Share story

Infrastructure

Netlify moves Edge Functions into its own network with faster invocations

Netlify moves Edge Functions into its own network with faster invocations Image: Primary
Netlify says its rebuilt Edge Functions infrastructure is serving production traffic through Firecracker MicroVMs inside its own edge network, replacing an external hosted execution service. The company reports median warm invocation times of approximately 5-6ms, down from 25-40ms on its previous infrastructure, and 47.4% faster p99 invocations. Requests now move from an edge node to a compute node within Netlify's network. Separate lightweight virtual machines isolate deployments, and snapshots allow functions to scale to zero and restart without loading an entire snapshot into memory. The migration requires no project changes and retains the same pricing. Cold invocations occur on about 1.2% of requests and take about 9ms on average.
Sources
Published by Tech & Business, a media brand covering technology and business. This story was sourced from netlify.com and reviewed by the T&B editorial agent team.
Back to Newswire
Keep reading
Full wire
Products Infrastructure
Products Infrastructure

HPE announces $1.2 billion Vultr order and raises networking outlook

Hewlett Packard Enterprise announced a $1.2 billion order from cloud company Vultr and raised its networking revenue forecast, Bloomberg reported. HPE now expects networking sales growth from the high teens to the low 20s in perce...

Infrastructure Products
Infrastructure Products

Accelevation raises $540 million in IPO priced below marketed range

Accelevation Holdings and its private equity backer raised $540 million in an initial public offering priced below its marketed range, Bloomberg reported. Shares of the data center infrastructure company fell 2.5% after the offeri...

Infrastructure AI
Infrastructure AI

CoreWeave opens Vera Rubin cloud access with Cognition in production

CoreWeave announced availability of NVIDIA Vera Rubin NVL72 systems on its cloud, with Cognition the first customer running production workloads, Nvidia said. Early-access customers can use the platform with Spectrum-X 102.4T Ethe...

Infrastructure AI
Infrastructure AI

Google opens cloud command tools to AI agents in public preview

Google Cloud has made a remote server for its command-line tools available in public preview, allowing AI agents to run cloud management and BigQuery commands without installing those tools locally. The server exposes operations f...

Infrastructure Security
Infrastructure Security

Databricks makes native IP analytics functions generally available

Databricks made built-in IP address functions generally available on Runtime 18.3 and above, allowing SQL users to parse, validate and join network data without custom libraries. The functions handle IPv4 and IPv6 addresses and ne...

Infrastructure AI
Infrastructure AI

Nvidia makes cuObject storage libraries generally available

Nvidia announced general availability of cuObject client and server libraries for accelerated access to object storage. The software uses remote direct memory access to transfer data without routing it through CPU-controlled memor...