faq @ voltage park

faq @ voltage park

Basic Responsibility and Crediting

Root Cause
Scheduled
Unscheduled
Customer / User
ie. driver failure, kernel panics, OS updates Because we give bare metal access to customers, there are many ways where a customer could break or freeze their hardware. This often requires a staff network engineer to resolve. Balance will not be credited for downtime in this case. Int18 will give “best effort” during business hours to resolve.
Int18 Supermicro, Nvidia, Dell Cisco, Juniper, etc
ie. Security updates, new deployments, repairing damaged hardware These are mostly important updates or fixes that help Int18 provide more secure, better services, to its customers. We need to ensure that our systems are tolerant to known attack vectors, and have important system tooling in place. Reasonable notice will be provided during business hours. Balance will be credited for over 12 hrs of downtime ex. 11 hrs 59 mins downtime = 0 credits 13 hrs 2 mins downtime = 13 hrs of credits
ie. System failures Int18 runs Enterprise software such as the NVIDIA DGX management console, as well as proprietary tools that help us monitor and ensure the health of the cluster. These run with minimal resource usage in isolated environments and will not impact customer workloads. Credited for over 12 hrs of downtime
Infra Vendors ie Datacenter Operator Internet Provider Power Providers
ie. Generator / substation maintenance There are important infrastructure routines that could cause potential downtime for customers. These include repair and maintenance of power generators or substations, HVAC systems, etc. These are vital for operator safety, and long term stability of all equipment. Reasonable notice will be provided during business hours. Credited for over 12 hrs of downtime
ie. Network or power failures Potential power outages due to extreme weather conditions, natural disasters like earth quakes, etc. Credited for over 12 hrs of downtime