Droplet killed by OOM on Hypervisor

December 30, 2018 494 views
DigitalOcean

Today my monitoring alerted me that the server on DO is down. Checking in the console I found the droplet in a stopped state. Started it back normally, DO support team replied in a ticket that droplet was killed by Out-of-memory on the hypervisor. Luckily this resulted in just a minor corruption of MySQL tables.
I’d like to know if this event is considered “ordinary” by DO team, as support tried to upsell optimized droplet to me as a “cure”. Anyone else experienced this?
I find this incident very disturbing, even more so because droplet was not started back automatically after being killed.

1 Answer

Hey there,

This is not “ordinary” behavior. What you describe was a hypervisor malfunction. I’ll provide more details in the ticket.

Eris

Have another answer? Share your knowledge.