Here at Hostinger we value our customers time and that’s why we are trying to be as efficient as possible. In past few weeks, I noticed an increase in some specific inquiries of our Customer Success department guys, and that was “Client is getting 503 error, maybe you can see what’s wrong?”. Hostinger is using CloudLinux for its vast farm of web hosting infrastructure and since our servers are not made of rubber, we have some safe limits in effect. As you probably already understood, clients are regularly reaching them, that’s the main source of that not very informative 503 error.
Our DevOps and Devs teams already have a plan to show which and when the limit was reached for the customer, but as I later was informed, this is not a priority right now. Well, I thought, how can we reduce this delay in answering the client why his website isn’t working. We already have the data from CloudLinux LVE module, it’s just a matter of how to show it to the Customer Success team.
I found this little plugin for our Slack bot which nicely integrates with Grafana metrics platform that we use regularly. I will not describe in detail the setup process, as it’s very well documented on their Github page. It took just a few steps to deploy the plugin via Chef and setup Amazon S3 bucket for storing the screenshots from Grafana.
Now, our Support guys and gals can access all limit metrics just writing one command to the Slack bot:
*yourbotname* graf db hostinger-lve-metrics
To sum up, now our client’s are getting the information faster as we eliminated the need to login to the server and check every case manually, Technical Support team also can take a breath and concentrate on other time-critical issues.