Mission Critical Linux website: Article / Services

Services

	Services, page 2

	mclx Friday 19 October 2007
\| Cluster technology \| Deployment Services \| Technical Support \| MCLX Cluster technology *Benefits* Proficiency in various cluster technologies Skilled professionals A solid foundation: the Linux operating system 7x24 technical support for cluster systems *Overview* Our company has been dealing with Linux-based cluster technologies since the very start. Although we found that so-called high availability clusters are sought after most, we have the professional knowledge and experience to build other types of clusters as well. Whether it is a high-performance technical, load balancing or SSI cluster system, our experts with many years of experience have the necessary competence to design and implement the cluster system meeting the requirements of our customers best. We also provide technical support for the clusters we deployed - if required, even with a 7x24 availability. *Proficiency in various cluster technologies* Our company has experience with the following well-known cluster types: High availability clusters In a high availability cluster (HA-cluster) several computers and peripheral devices are connected in a way that they form one system that stays operational even if some components fail (there is no 'single point of failure'). While earlier only vendor-dependent solutions were designed for the hardware of the given manufacturer, today it is possible to use Linux and standard PC components to build an HA-cluster. A typical HA-cluster includes the following components: two servers SCSI or Fiber Channel shared disk subsystem serial or Ethernet heartbeat channels (cluster communication interconnects) programmable power controllers to allow the servers to switch off the other one if necessary A KVM-switch or a serial terminal server to access the consoles HA clusters are generally used by our customers in those - e.g. banking or enterprise - environments where the continuous availability of the applications is very important and even a short outage may result in serious losses. High-performance technical clusters This cluster type - which is sometimes called Beowulf after one of the first such implementations - aims to realize supercomputer-like performance. These clusters join several computers to solve scientific/technical problems that require huge computing performance. They yield a very good solutions for those applications where the tasks can be processed well in parallel, e.g. for physical, mathematical or meteorological computations. The deployment of a Beowulf cluster usually consists of the following steps: selecting the hardware systems, implementing a Gigabit/Ethernet network installing and configuring the Linux distribution approved by the customer, tuning the kernel installing security settings and patches on the central machine building a scheduling systems according to the needs configuring the relevant process communication directories configuring the file server creating an image system based on MAC addresses creating workstation images according to the hardware configuration setting up boot configurations, TFTP and DHCP setting up the automatic management interface for client monitoring Load balancing clusters Load balancing clusters are ideal for web servers and electronic commerce applications. One of the most popular implementation is the Linux Virtual Server (LVS), which uses several servers to run the same application with static data or using the same database. Incoming requests are accepted by a separate load balancer server that forwards them to the web servers. Responses are sent directly to the clients. The web farm behind the load balancers can be changed dynamically, meaning web servers can be put in and taken out. SSI clustering A Single System Image (SSI) cluster is a system that is built of several physical computers, yet looks as only one for the applications it is running. Ideally, it can run applications written for non-cluster environments without any modifications. SSI clusters yield the benefits of both scalability and high availability. Its best known example under Linux is MOSIX (Multi-computer Operating System for UnIX). The various types of cluster technologies may be combined: for example, a comprehensive solution may include a load balancing web farm and the availability of its back-end DBMS is guaranteed by an HA cluster. *Skilled professionals* Our engineering team consists of experts with related professional university degrees. They have several years of experience in installing and configuring Linux-based cluster systems, also in solving the various related tasks and potential problems. A solid foundation: the Linux operating system Linux is a mature, solid, reliable multi-purpose operating system that performs well in any role of the enterprise IT infrastructure. That's why we focus on cluster technologies based on the Linux operating system. Under Linux a wide range of different cluster solutions are available, most of them in an open-source form that is free of charges. 7x24 technical support for cluster systems In order to further extend the safety guaranteed by high availability (HA) clusters, we provide - if needed, 7x24 - technical support for our deployed systems, in a construction that is most suitable for our customers. Technical support may mean availability, systems management or the combination of both. Previous Page << 1 2 3 4 5 >> Next Page

>> List articles in this category << Back to articles front page