Investigate root cause of db 500s during error spikes

[busbyk]

We are getting alerts from Vercel about error spike incidents. These look like attempted bot attacks that are failing to find anything vulnerable and we have concluded that these are safe to ignore.

The curious aspect to these error spikes is that our db calls eventually start to 500. We suspect that this is a safety mechanism from Turso but this needs further investigation.

Goal of this task: understand the mechanisms at play here.

Answer:

• Why is are db fetches throwing 500s during these error spikes?
• Re-confirm that these error spikes are "normal" events and that we can't do anything further to prevent them aside from maintaining a lack of vulnerabilities to be exploited
• See if we can establish differing notifications for error spike alerts that are harmless and ones that look different than what we've been seeing so we don't run into a boy who cried wolf situation where we ignore something different

[rchlfryn]

I was able to look into this a bit more and here is what I found:

Why is are db fetches throwing 500s during these error spikes?

We are getting 500s because a bot thinks we have a wordpress site and is searching for file vulnerability (See vercel logs)

Re-confirm that these error spikes are "normal" events and that we can't do anything further to prevent them aside from maintaining a lack of vulnerabilities to be exploited

I will make a PR to update the robots.txt file which is causing some of these 500s and keep this open to investigate further traffic. I also turned on bot protection logging