Cloudflare experienced another outage, causing prominent websites to go down. A solution appears to have been found, but the DNS provider is monitoring whether it is working properly. During a recent major outage, it became apparent that such a fix did not permanently restore the connection.
Cloudflare reports problems with the Cloudflare Dashboard and related APIs. During the previous outage, which lasted several hours last month, further investigation revealed that the Bot Management system received an update file that was too large. This was refreshed every five minutes, causing the problems to worsen worldwide as more nodes received the update.
This time, far fewer websites were affected, but LinkedIn, IKEA, and ourselves were among those impacted.
Temporary problem
The error message has now disappeared from Cloudflare’s status page. It therefore appears that this outage was short-lived. At the same time, regular maintenance is being carried out in various US regions.
Nevertheless, we remain cautious about the further consequences of such internet problems, especially since the previous major outage initially appeared to be short-lived. Websites were sometimes available, sometimes unavailable, and after a while, all websites using Cloudflare’s Bot Management system went down. However, this only happened some time after the first reports of the outage.
Read here about the major Cloudflare outage and what exactly went wrong.