I wanted to give a quick update on Crowdcube’s readiness for the upcoming fundraising round.
Rich goes into detail about what happened last time in his post but I wanted to share our thoughts ahead of this months raise.
Following the last round, our first move was to engage with Crowdcube’s engineering team and understand what had happened. As Rich said, the root cause was a missing index (something which helps databases work efficiently) and has now been fixed.
Addressing the previous issue isn’t enough though. To prevent something similar happening again Crowdcube have taken a more holistic approach to their testing. Testing individual parts of the system for an increased load is a good start but the only way to identify unknown unknowns - aka issues you can’t plan for in advance - is to simulate real-world behaviour. The engineering team at Crowdcube have been testing these real-world scenarios using extremely high load. They’ve been able to simulate traffic that’s a magnitude higher than our previous peak. Additionally, they’ve changed their product so it can downgrade more gracefully - meaning in the unlikely situation that things don’t work you won’t have any unwanted side effects.
In addition to this, the team are planning to host a war room. This means key members of Crowdcube’s engineering team will be monitoring their services and working to tweak things in real time should anything unexpected occur. We’ll also have a direct communication to the war room so we can provide instant feedback if you have any questions during the event.
For a more general look on the crowdfunding round - check out Adam’s post: Freetrade's crowdfunding on 25th June at 11am