Public IP service planned maintenance [Sunday, December 25, 2016]
Scheduled Maintenance Report for Ravello Systems
The scheduled maintenance has been completed.
Posted over 1 year ago. Dec 25, 2016 - 14:32 UTC
Please be advised that Ravello Systems will be performing service maintenance for the Public IP service. The maintenance is scheduled for Sunday, December 25, 2016, between 13.30 PM – 14.30 PM, UTC.

The maintenance will partially affect some running applications (see detailed description below). In order to completely avoid the impact, please apply the suggested changes in the mitigation section below.

Impact on Running Applications:
The maintenance will only affect applications that use the "Globally Available" US-West or US-East IP locations.
All other running applications will not be impacted.

How to determine if your application is using such IP:
IPs from the following ranges will be affected:
- In case your application is using a public/elastic IP from a different range - it should not be affected. (e.g : 104.X.X.X range or 54.X.X.X range)

During the 1 hour of maintenance affected applications' VMs may experience intermediate network connectivity issues and there will be up to two cycles of public IP address outage periods of up to 10 minutes each (inside this maintenance window).

During the public IP address outages, external access to VMs in the affected applications will be impacted as follows:
- It will not be possible to access VMs over Public IP addresses.
- Outbound access from VMs to the Internet will not be available.
- Attempts to perform the following operations in the specified locations/regions will fail:

Publishing new applications/VMs with Public or Elastic IP addresses from the US-West or US-East IP locations.
Starting VMs in those application.
Nonetheless, during the outage:
- Internal application communication (LAN) will be available using internal IP addresses/hostnames.
- Console access to VMs will be available.
- Other Regions and Other Elastic IP address locations will not be affected. (e.g : us-central-1 and "non Globally Available" EIP locations).

Optional Mitigations
It is possible to completely avoid the impact of the planned maintenance by taking one of the following measures:
- You may move the workload in advance to us-central-1 region In order to do so, simply take a Blueprint of the existing application and publish it specifically to Us-central-1 using Public IP (not elastic).
- You may avoid using Elastic IPs from the “Globally available” US-West/US-East locations. If you require an Elastic IP, you must publish you application on Us-central-1 and select EIP from US-Central-1.
- You may change the external access mode on all NICs in applications to Port-Forwarding for the duration of the maintenance.

General impact: Low.
Max expected impact time: 2x10m.

If you have any inquiries regarding this maintenance, please access
Posted over 1 year ago. Dec 20, 2016 - 21:13 UTC