The Scale of Our Systems

Many engineers arrive at Infobip from smaller companies, where they weren't accustomed to the scale Infobip operates at. What do we mean by that? Let's talk some numbers.

At the moment of writing this handbook, Infobip‘s underlying systems consist of:

• On-premise equipment installed in 35+ worldwide datacenters, which consist of:

• Over 1600 physical servers;

• Over 37.000 virtual machines;

• We are rapidly expanding into the cloud (AWS, Azure, Google), where applicable.

• All supported with over 1000 network equipment devices that keep things connected.

On those VMs, we run our microservice architecture, which consists of:

• Over 4.600 applications (service types),

• that, due to redundancy and architecture, run over 43.000 production application instances.

From a data perspective, we have:

• more than 700 relational database servers,

• with more than 9.000 databases,

• containing more than 600.000 objects (tables, Kafka topics, ES indexes),

• with a total of 600+ TB of working data size.

Our central internal data pipeline processes more than 800.000 messages per second as they pass through our Kafka clusters.

Our 100+ teams deploy application-level changes to production over 2.000 times a day, making us truly agile in terms of continuous delivery. The whole system processes:

• over 30.000 client messages per second,

• and transfers over 4.5 Gbps data between our datacenters.

From the client side, we have:

• More than 40.000 clients use our platform through our portal each day,

• and peaks of over 50.000 API requests per second,

• which in the end sends the data to 650+ mobile network operators that we directly connect to.

So those are the numbers to have in mind when we talk about Infobip's scale.