Since we have 300k+ happy users globally using our bots
We have to make sure that our bot runs stable and provides lowest down time as much as possible
To achieve that goal we’re using very fast and stable Azure VPS and heroku servers for deployment
Yes we are mainly using 2 virtual machines in our bot , it is similar to the concept of parallel computing
Here 1 question may come in your mind , ‘Why and how we are using 2 virtual machines for single bot?’.
For that We’d just created the separate repository carrying the functions of music player of our bot thus we’re able to deploy it in separate machine
It also helps main bot to run stable since we’d divided workload between 2 virtual machines
Database
:- To use 3 different libraries and store data for each modules in single database is possible but it is quite complex to do in real , also thousands of groups are being managed by out bots so it may be very expensive to store all data of bot in single database and it may lead buffer / downtimes in our bot
:- to fix that issue we’re using 3 different databases separately dedicated to the modules of 3 different libraries we’re using
Here is the list of databases we’re using
MongoDB
we’re mainly using MongoDB for some basic modules which aren’t based on PTB library coz PTB doesn’t supports MongoDB
PostgreSQL
Here PostgreSQL DB provided by elephantSQL is mainly being used in modules based on PTB ( Python-Telegram-Bot ) library
RedisDB
We’re just using redis DB for caching
Azure Virtual SSD
The Virtual data disk of Azure is being used in our 2ndary application, mainly used for playing music in voice chat
