Series of Facebook Games Managed by a Load-Resistant Platform

Itransition implemented support and maintenance for Facebook gaming platform.

Problem

Our customer ― Product Madness Ltd. ― is a US based innovative company specializing in Facebook gaming. Following its ambitious expansion strategy, the company planned to launch a number of new applications and to improve those already on the market, and Itransition team performed the task. The scope of work performed by our specialists is quite wide, ranging from casual games (e.g. farm-like, social quizzes, etc.) to gambling (e.g. online poker, blackjack, roulette). Another large-scale task completed by our team was to build an efficient software platform capable of managing applications and advertisements, handling accounts and billing, collecting and processing user statistics.

The developed host of Facebook games supported by a robust and scalable, load-resistant platform has a large amount of data to process.

100

million users for one installation

450

thousand active users daily

2.2

million unique users monthly

40

thousand requests per minute with average response time of 65 ms
120 GB of data store daily with further analytical processing of its monthly volumes

The Customer’s existing infrastructure comprises 10 applications. A minute of downtime can lead to thousands of dollars’ worth of losses for the Customer, which is why it is of crucial importance for their company to have professional support and maintenance to ensure system stability. Since the development part of the project was very successful, the Customer decided to entrust the support and maintenance of the project with Itransition support engineers.

Solution

Itransition team that works on the project consists of a Level 1 (L1) support engineer, Level 2 (L2) support engineer, and a project manager. They have certain responsibilities within their areas of expertise; overall, L1 support comprises 24/7 monitoring and incident escalation, whereas L2 support services are provided towards a list of specified cases:

L1 L2
Initial processing of monitoring alerts and customer requests Adding new app to monitoring, new hosts to Zabbix
Filtering and confirming real alerts, testing and proving incidents, escalation to the appropriate teams (whether it is L1 support engineer, L2 support engineer,  developer, or QA specialist) Faye restarting
Incidents escalation to third party providers (e.g. to a hosting provider if the problem is with the servers going down because of a blackout) Checking for high difference between master and slave, low free disk space, whether SQL slave is not running
Providing additional information about parameter trends (from monitoring) on request to the Customer or Itransition team working on the project Stunnel restarting
Assistance in incidents resolving on the Customer or Itransition team request Swap issue eliminating

Results

Thanks to in-depth technology and methodology expertise, Itransition support and maintenance specialists provide better end user experience making the Customer’s product run smoothly 24/7, without fail.