Taking AI from Pilot to Production
A field guide to the seven decisions that separate an AI experiment from a system your organization can operate.
Lire la noteNotes de terrain / Dernières nouvelles
Des notes d’ingénierie issues de l’exploitation d’infrastructures ouvertes : pannes, décisions de conception et travail upstream qui améliorent l’infrastructure ouverte.
Parcourir toutes les notesA field guide to the seven decisions that separate an AI experiment from a system your organization can operate.
Lire la noteCompare public, private, and hybrid cloud by cost, security, performance, compliance, operations, and workload fit to choose the right model for your workloads.
Lire la noteHow to choose between a consumption-priced API and dedicated capacity with some sample costs.
Lire la noteWe're always looking for new ways to improve the quality of the service that we deliver to our public cloud and managed private cloud customers.
We're always looking for new ways to improve the quality of the service that we deliver to our public cloud and managed private cloud customers. As networking demands increase, the use of a larger MTU allows virtual machines to transfer more volumes of data and use fewer packets which undoubtedly also introduces an improvement in throughput.
Over the past few months, we've been working hard at making internal changes to our networking infrastructure in order to be able to allow our users to have faster performance for their networking and we're happy to announce that starting now, any new private networks created in our public cloud will have an MTU of 9000, or the ability to have any MTU up until that size.
It's important to note that public internet traffic will continue to live at the standard MTU of 1500. However, internal traffic can go for much higher MTUs which can allow our customers to do more things like better throughput for overlays for Kubernetes clusters, avoiding mismatched MTUs with Docker daemons and other creative things we love to see our customers do.
Available on new networks
This change is already in effect and put in place for all of our regions and you should be able to see this reflected in any new networks created. If there are any existing networks that were created prior to this change, they will not have any modifications due to the fact that MTU must be matching everywhere and you would see packet drops if you had instances with different MTUs.
We hope that these improvements help make for a better experience inside our public cloud environment!
Choose from Atmosphere Cloud, Hosted, or On-Premise.
Simplify your cloud operations with our intuitive dashboard.
Run it yourself, tap our expert support, or opt for full remote operations.
Leverage Terraform, Ansible or APIs directly powered by OpenStack & Kubernetes