Status impact color legend
- Black impact: None
- Yellow impact: Minor
- Red impact: Major
- Blue impact: Maintenance
[Serverless Containers] - [nl-ams] - mandatory operation to prepare for new v1 API
Scheduled Maintenance
Report for Scaleway
Completed
The maintenance is finally done.
A few containers that were in "error" status before the operation might now be unreachable. This is because the latest deployed container revision wasn't working, and redeploying it failed. Users with such containers (error message starts with `Container is unable to start`) must redeploy a working version, or configure the probes so that container instances can start in due time.
Some other containers might now be in "error" status if the image has been deleted from the registry. These containers are not reachable anymore. Users with such containers (error message is `image was not found in container registry`) have to redeploy the container with an existing image.
Finally, on v1 containers, since probes (health checks) now can restart the container if it's not starting in due time, users have to configure the probes accordingly. This is a new feature, but a breaking change from v1beta1, so we will add it in the documentation here: https://www.scaleway.com/en/docs/serverless-containers/reference-content/v1-migration-guide/#breaking-changes. Sorry about any inconvenience.
We invite users facing other kind of issues to reach us through a support ticket that will be prioritized.
Thanks for your understanding.
A few containers that were in "error" status before the operation might now be unreachable. This is because the latest deployed container revision wasn't working, and redeploying it failed. Users with such containers (error message starts with `Container is unable to start`) must redeploy a working version, or configure the probes so that container instances can start in due time.
Some other containers might now be in "error" status if the image has been deleted from the registry. These containers are not reachable anymore. Users with such containers (error message is `image was not found in container registry`) have to redeploy the container with an existing image.
Finally, on v1 containers, since probes (health checks) now can restart the container if it's not starting in due time, users have to configure the probes accordingly. This is a new feature, but a breaking change from v1beta1, so we will add it in the documentation here: https://www.scaleway.com/en/docs/serverless-containers/reference-content/v1-migration-guide/#breaking-changes. Sorry about any inconvenience.
We invite users facing other kind of issues to reach us through a support ticket that will be prioritized.
Thanks for your understanding.
Update
We have completed half of the maintenance as of now. No major issues have occurred so far, except for a small hiccup with crons between 4:00 and 5:00 PM. A few crons were stuck in "upgrading" status during that time; however, they were triggered correctly despite the reported status. Therefore, no crons should have been missed.
If you have encountered issues with your crons, or your containers in general, in the nl-ams region today, feel free to create a support ticket linking this maintenance. We will look into it as a priority.
We will continue with the remaining containers tomorrow.
Thank you for your understanding.
If you have encountered issues with your crons, or your containers in general, in the nl-ams region today, feel free to create a support ticket linking this maintenance. We will look into it as a priority.
We will continue with the remaining containers tomorrow.
Thank you for your understanding.
In progress
Scheduled maintenance is currently in progress. We will provide updates as necessary.
Scheduled
On March 2nd and 3rd (between 8 AM and 5 PM UTC), we are going to perform on nl-ams a required operation to prepare the new v1 API arrival (available soon).
During this operation, changes will be performed to all Serverless Containers namespaces on nl-ams region.
While this operation is seamless most of the time, users may observe the following behaviors:
- Namespaces will pass temporarily in "upgrading" status. This may take up to 15 minutes, depending on the number of resources inside the namespace. During this time, all write operations on the namespace (update namespace, create container, update container, delete container, etc.) will be impossible. Read operations will work as usual.
- Container instances will be created. This is mandatory to ensure the containers are working properly. If a container fails to start, the "status" field will change to "error", and the "error_message" field will show the error, so users can fix it. During that process, users could see multiple instances of their container running, even if it's configured with max_scale = 1, and even if there is no incoming traffic. The Container will downscale as soon as the operation is over.
- Long running requests made during this operation may fail, as we roll out Container instances.
- Crons might be triggered multiple times during the operation. Crons with frequent schedules (such as every minute) might be more impacted.
Please note that during this process, unless the containers cannot start because of an applicative error, the containers will always be reachable.
This process is mandatory to prepare a seamless upgrade to the new v1 API coming soon, with new features. More details here: https://www.scaleway.com/en/docs/serverless-containers/reference-content/v1-migration-guide/
Thanks for your understanding and sorry about the inconvenience it may cause.
During this operation, changes will be performed to all Serverless Containers namespaces on nl-ams region.
While this operation is seamless most of the time, users may observe the following behaviors:
- Namespaces will pass temporarily in "upgrading" status. This may take up to 15 minutes, depending on the number of resources inside the namespace. During this time, all write operations on the namespace (update namespace, create container, update container, delete container, etc.) will be impossible. Read operations will work as usual.
- Container instances will be created. This is mandatory to ensure the containers are working properly. If a container fails to start, the "status" field will change to "error", and the "error_message" field will show the error, so users can fix it. During that process, users could see multiple instances of their container running, even if it's configured with max_scale = 1, and even if there is no incoming traffic. The Container will downscale as soon as the operation is over.
- Long running requests made during this operation may fail, as we roll out Container instances.
- Crons might be triggered multiple times during the operation. Crons with frequent schedules (such as every minute) might be more impacted.
Please note that during this process, unless the containers cannot start because of an applicative error, the containers will always be reachable.
This process is mandatory to prepare a seamless upgrade to the new v1 API coming soon, with new features. More details here: https://www.scaleway.com/en/docs/serverless-containers/reference-content/v1-migration-guide/
Thanks for your understanding and sorry about the inconvenience it may cause.
This scheduled maintenance affected: Elements - AZ (nl-ams-1, nl-ams-2, nl-ams-3) and Elements - Products (Serverless Functions and Containers).