Zabbix at GTD Group, Case Study

“As one of the leading service providers in Chile, our key virtues drive us to be agile, reliable and caring for our customers’ experience and needs. Operating in the Internet, Telephony and TV sectors implies availability and consistency of everything we do. Being able to address all that is why we chose Zabbix.” – says Joel Urtubia Ugarte, senior network intelligence engineer at GTD Group.

GTD group currently caters to more than:

  • 250.000 internet customers;
  • 160.000 TV pay customers;
  • 250.000 telephony customers.
Objective:

GTD group needed a scalable solution capable of monitoring the entirety of their infrastructure out of the box, as well as reliable community or commercial support.

Requirements:

Taking into account their number of customers, GTD group needed to monitor a vast array of difference devices on their network. As most of them support SNMP, achieving this with Zabbix was trivial. In addition to that, it was essential to keep and effectively manage historical data collected from all the devices. With that in mind, further requirements implied support of faster DB engines, such as Elasticsearch, and powerful API as well.

Approach:

Having previously used Cacti and HP OpenView (now HP BTO Software), the first noticeable difference in approach was the ease of the actual setup process with Zabbix, thanks in part to the tool’s built-in features and documentation, as well as in-house knowledge of GTD group’s Engineering Manager, Ciro Vera, who worked with the tool previously.
According to Joel Urtubia Ugarte, senior network intelligence engineer at GTD group, the one challenge their team faced at some point is the learning curve when it comes to tuning the Zabbix server for a higher workload. This, however, was quickly resolved with official Zabbix Certified Specialist and Certified Professional training.

Business outcomes:

“You can monitor any device, that’s is the greatest feature that Zabbix has” – Joel Urtubia Ugarte, senior network intelligence engineer.

  • Full visibility of the network and customer devices, allowing to know of any outage before the end-user and take necessary actions;
  • Ability to monitor almost any network device out of the box and have the data visualized in form of graphs and maps;
  • Flexible triggers to be notified on root causes of the problem;
  • Improved performance and stability of monitoring;
  • Very low TCO compared to previously used monitoring tools.

With Zabbix setup and running since 2010, the team lead by Ciro Vera managed to effortlessly monitor well over 100’000 devices all with a single Zabbix server, thus being able to guarantee quality services to its customers.

Zabbix stats:

Architecture: 1 Zabbix server with Elastisearch as the choice of DB
Number of hosts: ~120’000
Number of items: ~630’000
Number of triggers: ~320’000
NVPS: ~4’500

Jekaterina Petruhina

Author: Jekaterina Petruhina

Marketing Specialist at Zabbix

Leave a Reply