Scaling and cluster for Magento

Can you imagine an online store serving 81 000 visitors per day at data-center costs 15$ an hour? Delivered!

Our company can practice automated scaling with any data-center (even your own).

Automated scaling provides availability and web-site speed at consistent high level. Resources scaling provides both increase and decrease — It depends on required conditions. We automatically gather data on web-site load and in real time regulate the number of servers dealing with your web-site. Automated scaling — is a unique and reliable tool for supporting web-site availability at different load peaks (New Year, after TV advertizing etc.)

How we achieve fixed response time at any load?
We keep configurations
in one place
We use GIT for tracking
configurations changes
We centrally manage
virtual servers
Puppet - for configurations
management
We provide real time
productivity control
NewRelic/Zabbix - for tracking
server loads and response time
We have advanced safe containers
for quick server deployment
Docker
DBMS scaling
Perconna XtraDB Cluster
We spare your servers the unnecessary work
Varnish, Pagespeed - aggressive
caching

Autoscaling

The number of servers changes depending upon the number of users

Autoscaling helps you use the cloud resources with a maximum efficiency and save your money by increasing resource levels under the increased load and reducing them when the load is reduced.

When Autoscaling is on, it monitors the occupied server resources. Changes of resources available and a step of such changes depend on the regulations set by the user and the autoscale goal. Some projects need a possibility of quick scaling while the others need to save more money.

At any data-center

or using
your
own equipment

We can use any data center in Russia or even use your own servers for autoscaling.

!Some server platforms do not support Autoscaling. We will check every data center for compatibility by your request (free of charge).

No problems during pre-New Year days or large-scale promotion campaigns

SLA

An agreement under which we guarantee you high-quality services

SLA (Service Level Agreement) is used in corporations for regulating intra-departmental relations and moreover it's a principal tool of continuous evaluation and quality management of outsourcing services provided by a dedicated organization.

To the data-center To the client

Archive copies anywhere

Data backup is needed for quick and cheap data recovery (documents, programs, settings etc.) in case a production copy is lost by any reason.

Case study at 10$/hour

The Real case

The Real challenge

In 2012 client had more than 6000 orders during Black friday period. Our task was to implement server infrastructure ready to handle even more orders during 29.11.2013 – 02.12.2013.

The Real solution

A customized server infrastructure using Puppet/Docket has been developed by KT team on Magento eCommerce platform with a separate Varnish server, a separate Backend server, multiple Frontend servers and separate servers for sessions, caching and data storage. Frontend servers were increased from 4 to 18 during the high loads.

The Result:
2 to 5
orders
per minute
1500
visitors simultaneously
54 000
unique visitors

Still not enought?

SQL Master-Master
CDN
Varnish
Aggressive caching
NewRelic Monitoring
24/7 monitoring and support

Optimise your resources usage

Only for 11499$


Please note:

  • Autoscaling solves a performance problem: it's just as much as you need;
  • Autoscaling saves your resources: you pay just for the servers used at the moment;
  • Autoscaling is versatile and can be used with any web-based software (if allowed by your product license);
  • Komplizierte Technologie can make Autoscaling at any data center and even at your own servers (Hyper-V. ESX. KVM. EC2. CROC. Caravan. Everywhere*).
  • We will make a fully documented and described product with configuration files in a version check system. Your technical experts will appreciate it!
  • In case when a sharp load increase is planned (x20) we may increase a number of servers in advance.

* — technical capability is defined individually

Quality First

The architect, who took part in Magento development. One architect for every three developers. Mandatory Code Review. QA team. KPI, that places quality in the first place. In short, we really tried to assemble a team of people who care how everything works.