Infrastructure Maintenance
This page describes the different types of maintenance events and the associated communications sent by 3DS OUTSCALE.
To ensure a transparent process and allow you to plan ahead, a communication is sent to impacted customers at key stages of each planned maintenance:
-
Two weeks before maintenance (depending on the event), an initial notification is sent to inform customers of the upcoming maintenance, detailing its nature, the potentially affected services, and the estimated duration of the intervention.
-
One week before maintenance (depending on the event), a reminder is sent to remind customers of the date of the intervention.
-
At the beginning of maintenance, a notification is sent to specify the details of the intervention and any temporary disruption to services.
-
At the end of maintenance, a communication confirms that the intervention has ended and that service is fully restored, completing the follow-up.
|
To receive notifications, ensure that the email address associated with your account is valid. For more information, see Modifying Your Personal Information and Password. |
Scheduled Events
Scheduled events are planned maintenance tasks that can affect one or more of your virtual machines (VMs). A scheduled event can have one of the following codes:
-
system-reboot: VMs will be affected by software upgrade or hardware maintenance. -
system-maintenance: VMs will be affected by network or power maintenance, and may be temporarily unavailable.
To know which VMs are affected by a scheduled event, see Getting Information About Your VMs > Getting Information About the States of Your VMs.
TINA Upgrade
When a new version of our Cloud orchestrator TINA is available, maintenance is planned to install it in each Region of the OUTSCALE infrastructure.
These maintenance tasks are planned in the following order:
-
us-west-1
-
ap-northeast-1
-
us-east-2
-
eu-west-2
-
cloudgouv-eu-west-1
A TINA upgrade takes two hours for each Region. During this maintenance, the APIs are unavailable for 30 minutes.
|
This downtime causes an HTTP 503 code (Service Unavailable) when requesting the API and when connecting to Cockpit. |
An email is sent to impacted customers:
-
(cloudgouv-eu-west-1 only) Two weeks before maintenance
-
One week before maintenance
-
At the beginning of maintenance
-
At the end of maintenance
Hypervisor Maintenance
To keep the infrastructure up-to-date and ensure top-quality service, we regularly plan hardware maintenance on hypervisors in all our Regions.
Impacted customers automatically receive a support ticket two weeks before maintenance, including the list of their impacted VMs.
We strongly recommend stopping and starting those VMs before the maintenance date. This automatically transfers them to another hypervisor that will not be affected by maintenance, in the same Region. You can transfer your VMs as soon as you receive the maintenance notification email. For more information, see Stopping and Starting VMs.
|
If you do not transfer your VMs and they are still in the |