This blog post will cover a very important aspect of achieving the reliability you need: Operations – this means the day to day activities and arrangements you need to do to ensure that your device lab is able to operate 24/7 and it will keep producing repeatable results every day and night. Here are some of the best practices we have learned while operating Bitbar Testing:
This is very simple best practice but still many do not follow this and as a consequence end up chasing their own tail while troubleshooting complex issues. In its simplicity, many of larger enterprise setups are equipped with hundreds of devices and this requires 24/7 attention from the team. Operation teams should document all bits and pieces that happen to the environment, devices and test runs. By doing great homework with docs, the team will it easier to maintain and upgrade the environment when needed. Want to learn all of the tips of building in-house device lab? Download our ebook.
Keep your devices cool
Everything is about power nowadays. It has quickly become the most important things for all electronic devices. And that power and its consumption is the most important thing to contribute the temperature of the system. Even smaller changes in temperature can mean that lots of devices are malfunctioning (this should never be underestimated!) and go offline. When you set up the device pool, please make sure you let enough space for those devices to “breath”. If you stack them like in the picture below, you won’t see long-lasting lab – it will quickly put you in a fire, literally.
Monitor the charging of devices
You naturally need to make sure your devices get enough power to perform. Once those devices are 24/7 plugged in, maybe surprisingly, those devices need to be constantly monitored if they get enough power. The battery as hardware component doesn’t always allow full charging and easily gives up when it is constantly kept in charging mode. Make sure you monitor those devices!
Use staging environment to validate any configuration changes
Like with any software development, staging environment can help you a lot of verifying even smaller changes in the critical environment. In staging, you check that changes work as expected and validate all new configurations in that context. It is also a good idea to roll in new hardware (even devices) through staging if devices are ‘exotic’.
For a complete picture, you can watch our webinar recording about building large scale in-house test lab:
Standardize your test lab hardware as much as possible
It is always a good idea to use the same hardware when setting up or configuring a larger in-house lab. This means that if you use certain device setup for servers, it makes your operational team live easier the very same hardware is constantly used across the environment. Those devices naturally differ, but then again, you can use the same ways to connect, control and maintain those.
Find the maximum number of devices that can be connected to each server. Generally, and from our experience, the typical number of devices connected to one server for robustness is 15-20. Sometimes connecting devices in different USB slots in the same server may give much better stability. So it definitely pays back to find out what USB ports work the best. Also, minor trick but very important in case of huge setups: use color-coded USB cables and you know right away in what devices do they connect with.
As setting up is just one part of the story, maintaining and running operational point of view these labs is even more work and requires 24/7 attention. From our experience, the larger the environment is, the more challenging it will be to be built.
If you are interested to build your own lab and want to know all the best practices and recommended devices, hardware and any other aspects of it, please do not hesitate to contact us for more information.