The London furnace got the better of a Google Cloud data center on Tuesday, still disrupted twenty-four hours later, due to a cooling-related failure.
On Tuesday, the thermometer panicked and caused a major outage at Google Cloud’s London data center. One of the buildings that houses part of the capacity of the “europe-west2-a” zone suffered an incident, which was only partially resolved. It must be said that yesterday, it was more than 40 ° C in London, the English capital suffered an unprecedented heat wave, or almost. Fortunately, it has now passed, with a drop in temperature of between 10 and 15 degrees 24 hours later.
A cooling-related data center failure
On Tuesday, July 19 at 1:13 a.m. local time, the Google Cloud Services Dashboard reported an incident in its London region (the data center, therefore). He thus referred to a cooling failure of one of its on-site buildings.
The failure then caused a partial shutdown of the data center’s capacity, ” resulting in VM shutdowns for a small group of customers “. Instead, Google Cloud techs shut down part of the zone and limit Google Compute Engine launches (an infrastructure offering that allows enterprise customers to launch workloads on Google hardware).
Google reported around 12 p.m. Tuesday that the issue was largely resolved for all customers affected by the overnight outage, with the cooling system restored in the data center. Londoner.
Services still subject to disruptions, 24 hours later
But the Google Cloud Service Health console tells us of the interruption, again on Wednesday, of several Cloud services in the region. This is particularly the case with API Gateway, which allows its users to develop, deploy and secure APIs using a fully managed gateway. The App Engine service, a fully managed serverless platform for developing and hosting web applications, also continues to struggle. Usually, it is supposed to allow you to choose from different languages, libraries and frameworks (Go, PHP, Java, Python, Node.js, etc.) to develop your applications.
This is also the case for Google Cloud SQL, GC Storage, Google Kubernetes Engine or Persistent Disk services (already touched the day before and which helps to provide high performance for virtual machine instances). Cloud Tasks, the service for asynchronously executing tasks outside of a user request, is still experiencing disruptions too.
These products are currently experiencing error rates, latencies or quite simply unavailability of service, still linked to the cooling incident the day before. It is not yet known when these will be fully restored. Note that another major cloud player, Oracle, is also experiencing some heat-related disruptions in London.
Source : Google Cloud Service Health
10