[April 6, 2018]
Our uptime monitoring network has suffered a partial outage, due to DNS resolving issues from CloudFlare. Details below:
- 19:20 UTC we started noticing degraded DNS performance in our London and Frankfurt uptime monitoring locations.
Upon investigation we noticed that CloudFlare hostnames were not resolving properly in these two locations, and thus being reported as failing by our uptime monitoring nodes in London and Frankfurt.
- 19:38 UTC we shut down the London and Frankfurt uptime monitoring locations, to prevent further false positives from being reported by these nodes in regards to CloudFlare hostnames/websites, and we keep a close eye on the situation as it develops.
- 19:54 UTC CloudFlare acknowledges the issue and starts investigating: https://www.cloudflarestatus.com/incidents/9ddg7t9y9mh8
- 20:25 UTC we notice that CloudFlare DNS queries start resolving again, and our London and Frankfurt nodes are brought back online into the uptime monitoring network.
During this DNS outage our platform has issued Uptime Monitoring false positives for the following scenario (must have fulfilled all conditions listed below):
- client is monitoring the uptime status of a hostname/website which is under CloudFlare nameservers
- client has selected London and Frankfurt as their monitoring locations
- client has selected only 3 monitoring locations
Just a small number of our uptime monitors have been affected by this issue, and have received false positive downtime notifications within the time interval described above.