How can StackHPC help you?
StackHPC offers a variety of services both to organisations running clouds and organisations using multiple clouds to run HPC and AI workloads.
StackHPC has particular expertise and experience with dynamic infrastructure that provides agility without sacrificing the performance demanded by HPC and AI workloads.
For example, StackHPC uses LOKI (Linux OpenStack Kubernetes Infrastructure) to help extract the most value from their investments in infrastructure and people.
Creating a Private Cloud?
Looking to move away from silos of bespoke infrastructure towards a dynamic pool of reconfigurable infrastructure?
Want to start consuming local infrastructure via industry standard infrastructure-as-code tooling like Ansible, Packer, Terraform, and Kubernetes?
Want to use that same tooling for both virtual machines and bare metal servers?
Maybe you have heard about using Ceph to provide object, block and file storage?
Or you want to incorporate your chosen storage vendor's solution into a cloud-native environment?
StackHPC can help you adopt OpenStack using a combination of our well tested recipes and hard won experience across HPC, AI and cloud.
StackHPC have experience assisting with hardware procurements in an independent manner to ensure you specify what is best needed for your target workloads, and help you understand better the tradeoffs involved.
Support for your OpenStack Cloud?
StackHPC provides support for Kayobe based OpenStack clouds, often with integrated Ceph storage.
StackHPC helps customers sustain their investments keeping their cloud secure, on a supported and verified release and following latest good practices.
StackHPC provides assistance expanding and refreshing client hardware.
StackHPC helps clients exploit their cloud infrastructure to the fullest extent through cloud-native compute platforms and onboarding new workloads.
Training for OpenStack Operators?
StackHPC’s policy is to develop through training a greater degree of independence, reducing a client's reliance upon and cost of vendor support.
StackHPC provides HIIT Training and Workshops to build up the local skill set, capabilities and confidence in order to help teams tasked with OpenStack and Ceph administration become effective and self-sustaining as quickly as possible.
Help to Optimise your Workloads?
Do you have intensive workloads that need RDMA networking, access to high performance shared file systems and maybe GPUs?
Some of our clouds have Top500 and Green500 supercomputers as one of their workloads!
StackHPC’s speciality is helping people get the most out of their cloud and related investments. StackHPC collaborates to deeply understand what workloads are important and how best to meet all those needs using either your OpenStack cloud, and/or an external cloud like Azure, GCE and AWS.
Do your users need self-service applications? Do you want to offer pre-paid cloud credits, such that your users can reserve capacity when they need it, rather than squatting on those resources? Contact us about how the Azimuth project might help you!
Need Help with your OpenStack cloud?
Have you built your own OpenStack cloud, but would like advice on how to improve or fix some specific pain points?
StackHPC’s extensive experience working with OpenStack can be used to help get you and your cloud happy and productive.
StackHPC has adopted client-deployed OpenStack clouds by bringing them under the control of OpenStack Kayobe, and then upgrading them to a supported OpenStack release. StackHPC can also help you address underlying architectural problems in subsequent hardware refresh cycles.
Need Development Work to Integrate with OpenStack?
Open infrastructure supports a broad ecosystem of hardware, but further development may occasionally be required outside of the standard OpenStack and related services.
This may involve, for example, integration of open infrastructure services with specific data centre infrastructure, such as monitoring, logging and security platforms.
The HPC and AI domains are prominent innovators of processor, accellerator, network and storage technologies - all of which benefit from effective management in cloud infrastrucure.
Our API-driven, cloud-native and open standards approach ensures that key infrastructure can be integrated with full automation and maintenance.