Status impact color legend
- Black impact: None
- Yellow impact: Minor
- Red impact: Major
- Blue impact: Maintenance
[Serverless Functions/Containers] [fr-par] infrastructure maintenance (networking)
Scheduled Maintenance
Report for Scaleway
Completed
The scheduled maintenance has been completed.
Update
The last phase of workloads relocation is now complete. We are going to finalize the maintenance, but no other impacts are expected for these last steps.
Update
Main nodes have been updated. We will now proceed to the final relocation where users and system workloads will be moved back from temporary nodes to the main updated nodes. We will keep you informed once it is done. During that relocation, all running containers/functions instances are expected to restart once again. Possible impacts are: HTTP errors (403, 5xx) when calling functions/containers, requests taking longer than expected or timeouts, a few seconds of metrics loss in cockpit.
Update
The first phase of workloads relocation is now complete. We are going to update empty nodes, and then will perform another relocation to migrate workloads to updated nodes. We will keep you informed when this new relocation begins.
Update
503 errors mentioned in the last message are now resolved. This was caused by a side-effect of the maintenance, where some system components failed to restart correctly. It happened between 08:01 UTC and 08:31 UTC. Sorry for the inconvenience.
Update
One third of functions/containers are receiving HTTP errors 503 ("upstream connect error" and "connection failure"). We are on it.
Update
Users can also experience issues to deploy containers/functions (pending for a while, or ending up in error after some time). We are on it.
Update
In order to update our nodes configuration, we are now going to relocate users and system workloads to temporary nodes. During that relocation, all running containers/functions instances are expected to restart once. Possible impacts are: HTTP errors (403, 5xx) when calling functions/containers, requests taking longer than expected or timeouts, a few seconds of metrics loss in cockpit.
Update
Network services have been restarted. We are continuing the maintenance, we will keep you updated for next steps.
Update
We are restarting our network services. During around 15 minutes (we will keep you updated once it's done), possible impacts are: HTTP errors (403, 5xx) when calling functions/containers, network temporary issues, DNS resolution issues, requests taking longer than expected or timeouts.
In progress
Scheduled maintenance is currently in progress. We will provide updates as necessary.
Scheduled
A maintenance will be performed on our fr-par infrastructure for both Serverless Functions and Containers products.
Networking configuration will be updated for a long-term stability improvement. As such, instances hosting the users workloads (functions/containers) will need to restart to reload the new configuration. Some hiccups can be experienced (see "impacts" section below); we apologize for any inconvenience.
Impacts:
During the maintenance, clients might encounter the following disturbances:
- temporary network flapping: at the beginning of the maintenance, during around 5 minutes while network is reconfigured
- a few transient 403 and 5xx errors: mostly at the beginning of the maintenance, but also doing instances restart, while system components are restarted
- functions/containers unexpected restarts: as we restart underlying instances, running functions/containers will need to be relocated to different instances. Depending on the functions/containers configuration (min scale or how they handle termination), some 5xx can also be experienced.
Networking configuration will be updated for a long-term stability improvement. As such, instances hosting the users workloads (functions/containers) will need to restart to reload the new configuration. Some hiccups can be experienced (see "impacts" section below); we apologize for any inconvenience.
Impacts:
During the maintenance, clients might encounter the following disturbances:
- temporary network flapping: at the beginning of the maintenance, during around 5 minutes while network is reconfigured
- a few transient 403 and 5xx errors: mostly at the beginning of the maintenance, but also doing instances restart, while system components are restarted
- functions/containers unexpected restarts: as we restart underlying instances, running functions/containers will need to be relocated to different instances. Depending on the functions/containers configuration (min scale or how they handle termination), some 5xx can also be experienced.
This scheduled maintenance affected: Elements - Products (Serverless Functions).