Customers Area
Our Installations
For over ten years, we have consolidated installations of interconnected Linux physical machines with a distributed filesystem and a highly efficient control mechanism. Our infrastructure is based on Pacemaker, a robust cluster management software, which we tailor to meet customer needs. The software can be configured in bare metal to manage infrastructure services such as storage or HA firewalling, or in association with virtualization software like KVM, either plain or managed by VM control software such as ProxMox or similar. The figure below shows a typical Linux cluster with three physical nodes (which can easily be scaled up to dozens of nodes as needed), featuring a distributed filesystem—Ceph in our case—and a cluster manager, Pacemaker, to handle infrastructure services. The physical system is redundant and allows operation even in case of a complete node failure. It is scalable: increasing the number of physical nodes enhances failure resistance. The distributed filesystem ensures file and resource replication across different local partitions of various disks. This solution provides significant savings as it does not require a fiber-optic SAN and allows excellent horizontal scalability as the number of nodes increases (which is impossible with a SAN, for example). Moreover, the distributed filesystem enables hot service migration between nodes, whether they are standalone server services or virtual machines. From a clustering perspective, these resources are perfectly identical and can be moved from one node to another within seconds without service disruption (or packet loss).
Not only Virtual Machines
On this infrastructure, in addition to full-virtualization virtual machines, we can also install container management systems such as Kubernetes or OKD (or OpenShift) depending on customer needs—whether they wish to create a private cloud hosting hundreds of web applications rather than a virtual machine farm, or even both. From our perspective, once the core infrastructure is in place, it is simply a matter of customer preference, to which we adapt accordingly.
Deployable Services
On infrastructure services, whether virtual machines or container-based, we can deploy large-scale services such as video servers or large mail servers with thousands or even hundreds of thousands of accounts. Specifically, we offer installations of:
- High-traffic web servers, application servers, or API servers
- Distributed storage or dedicated video servers
- Mail servers
- Distributed Single Sign-On for authentication services
Commercial Cloud
If our solution does not meet customer needs or if they prefer to use a paid infrastructure such as AWS, Azure, IBM Cloud, or Google Cloud, we can still install and support intermediate services without any issues. We also offer support for migrating on-premise technology, such as VMware nodes, to the cloud of choice, ensuring minimal service disruption for the customer.
Monitoring
All our systems come with at least two monitoring mechanisms: Prometheus/Grafana and Nagios. The first provides real-time insights into the status of core services and proactively identifies behaviors that could lead to potential issues. The latter is a more "classic" yet highly robust control system that alerts the administrator immediately when an issue arises. We tend to use both for maximum security. Naturally, we are ready to install any monitoring system the customer deems necessary for their needs.
