« Back to all articles

Nearing the end: the final tickets

Written by Yi Ming Yung on 16th June 2016

My internship at the shack full of nerds called Devhouse Spindle is nearing its end. I have nearly spent twenty weeks with these lovely nerds whilst working together. The nerds in question wanted me to research and develop a monitoring dashboard to improve the visibility of the status of the VoIPGRID telephony platform. The status needs to be evident during outages, of course, but also when everything is operational. By having the status evident at all times, we want to achieve a higher level of transparency for our platform.

What have I done?

During my twenty weeks at Spindle, I have done research amongst our partners to determine what information is valuable on a monitoring dashboard for theVoIPGRID platform. After concluding the research and establishing the requirements for our monitoring dashboard, I have decided to use statuspage.io for visualization and develop middleware to automate this process.

Requirements

Many partners requested to see the status of the different servers of the VoIPGRID platform. By having these, partners could potentially perceive the status of the platform as a whole. In reality, the status of individual servers doesn’t say a lot about the platform because VoIPGRID has many fail-overs and one server being down won’t have a big impact on the performance. Our experts offered a more suitable solution, displaying the status on a higher level. Instead of displaying the servers that make up our services, we will display the status of the services themselves. We will be displaying the main services of VoIPGRID, namely:

  • Telephony
  • The web portal
  • Fax
  • Click-to-dial
  • The Vialer app

Pasted image at 2016_06_16 03_09 PM

This way, users don’t have to assume things and the user will be able to see a concrete status. Other requirements are latency, uptime and response time metrics. These metrics will be visualized using graphs on our status page. The initial setup of statuspage.io is shown below. This setup still requires the magic touch of our design nerds.

Screen Shot 2016-06-13 at 11.37.43

Status middleware

To make the status page reliable, I have developed middleware to automate the status of our services. The middleware will form a bridge between our internal monitoring system and statuspage.io. This means that when we detect an outage on one of our services, statuspage.io will be automatically updated. Right now, the services web portal, click-to-dial and Vialer app already have accurate monitoring. The services telephony and fax are being worked on to make an automatic and reliable monitoring. When this is done, the status page will be fully automated and will display the most real time and accurate status we can provide.

Conclusion

The monitoring dashboard / status page is basically done and now needs some testing. It is also waiting for the telephony and fax monitoring to be ready. When these are done, we will place the status middleware and statuspage.io in production. Further improvements in the future could allow the middleware to serve multiple status pages and allow partners to use their own statuspage.io with their own branding.

Last words

In my final weeks, I wrote my thesis, submitted it and got a 7,5. This means I will be graduating next week as Bachelor of ICT! I want to thank all the partners that took the time to help me with my research and my soon-to-be colleagues for their support. I am honored to become one of the supernerds in the near future.

thumb__MG_7860_1024

Your thoughts

  • Written by E on 27th July 2016

    Well done Yi Ming! Great decision to show the status of the services, people don’t care what’s the status of the servers that deliver these. Regards Eimert

  • Written by Yi Ming Yung on 17th August 2016

    Thank you! It is good to hear that people agree with this vision.

Devhouse Spindle