August 22 - September 08, 2018 - All regions

Last Update: 2018-09-04 16:00 UTC: Update special notice for Streaming Analytics and Geospatial Analytics service

Summary

In order to address an urgent, industry-wide security advisory, all virtual machines in the world-wide IBM Cloud data centers will need to be patched over the next several days, between August 22nd and September 8th. Due to the nature of this vulnerability and the components which are affected, we are not able to mitigate this potential vulnerability via Hot Patching and Cloud Hosts reboots are required. The patches/restarts will be staggered to minimize the overall impact to IBM Cloud users.

All start times are approximate and may be revised as we get closer to the event. Each maintenance period will last approximately 120 minutes. In many cases, the duration of each update may be less than the full maintenance window. Further detailed schedule information, as well as updates during the week, will be appended at the bottom of this page.

The IBM Cloud Engineering team will take all reasonable steps to minimize the disruption to the overall platform and its various components. We will closely monitor the progress and provide periodic general updates on this page, as necessary.

Cloud Foundry applications

In Public regions, because they are large and span multiple pods, we expect minimal disruption to the overall platform.

Every Cloud Foundry-based application instance will be automatically restarted at least once. To minimize disruption, we recommend using at least 3 instances for critical Cloud Foundry-based applications.

We strongly suggest that you take advantage of multi-region deployment of critical Cloud Foundry-based applications, where applicable, to minimize the disruption to your production applications.

Kubernetes service

Customers are responsible for managing their IBM Kubernetes Service worker machines.

Customers using the IBM Cloud Kubernetes Service are advised to make use of Multi-Zone Region capabilities where available (us-south, us-east, eu-gb, eu-de) in order to maintain optimal workload availability during the reboots. During the reboot cycles, customers may see periods of time where their Kubernetes masters are unavailable for a short period of time. This is expected and should recover automatically. It is important to note that the Kubernetes data plane is not dependent on the control plane. While the master cannot be called to create, update, or delete resources during an outage, all applications and networking function are unaffected during this outage.

Customers with mission critical applications should consider the use of multiple regions. See the documentation for more information.

Note: Your IBM Cloud Kubernetes Service bare metal worker nodes will not be affected by this maintenance. The Kubernetes master for these workers will still see an interruption in service.

IBM Cloud Dedicated regions

In individual Dedicated regions that typically are contained within a single pod, there will be a disruption in service to the environment which will include the following impacts:

  • IBM Cloud login and application management functions, such as staging, starting, stopping, and Services management might be unavailable.
  • Running applications might be temporarily unreachable.

If you have a Dedicated IBM Cloud region, you may contact your IBM Client Success Manager, Technical Account Manager, or IBM Advocate for more information.

Services

Many of the services that IBM Cloud applications are bound to may be temporarily unavailable as their data center servers are restarted or patched. Most of these service restarts will take place during the same maintenance windows listed below. However, some services might be affected at times outside these windows by other data center dependencies. See the table below for additional details about specific services.

Service or Component Name Maintenance Windows Notes
Watson Studio - Modeler Flow

Watson Machine Learning

EU-GB: 2018-09-07, 11:00 UTC - 2018-09-08, 01:00 UTC
There will be a brief outage for the Watson Studio Service - Modeler Flow tool and the Watson Machine Learning service in the United Kingdom public environment. This outage will occur within the 2 hour SL reboot window.
Watson Retrieve and Rank 2018-09-05, 05:00 UTC There will be a brief outage during the DAL06 POD 2 reboots. This 15-30 minute outage will occur within the 2 hour reboot window as the 2 devices housing the Ranker service are rebooted.
Streaming Analytics - VM Plans
  • US South: September 5, 2018 - 1:00 - 5:00 UTC
  • EU-GB: September 6, 2018 - 21:00 - 23:00 UTC
During this time, your Streaming Analytics Service will be stopped and unavailable. Once the service is back up, you must restart your Streaming Analytics service instances and resubmit your applications.
Streaming Analytics - Container Plans
  • US South: September 5, 2018 - 7:00 - 9:00 UTC
During this maintenance period, there will be periods of disruption to instances/applications of container based plans.
Geospatial Analytics
  • US South: September 5, 2018 - 1:00 - 5:00 UTC
  • EU-GB: September 6, 2018 - 21:00 - 23:00 UTC
During this maintenance process, all instances of the Geospatial Analytics service will be stopped and the service will be unavailable.

Schedule per DataCenter and POD

August 22
MEX01 POD 1 05:00 UTC

August 24
DAL10 POD 4 01:00 UTC FRA05 POD 1 03:00 UTC DAL08 POD 1 05:00 UTC DAL08 POD 2 05:00 UTC SEA01 POD 2 09:00 UTC TOK02 POD 1 15:00 UTC AMS01 POD 2 19:00 UTC OSL01 POD 1 21:00 UTC LON06 POD 1 23:00 UTC

August 25
FRA04 POD 1 01:00 UTC DAL01 POD 4 03:00 UTC DAL01 POD 5 03:00 UTC DAL06 POD 1 05:00 UTC SAO01 POD 1 07:30 UTC SEA01 POD 1 07:00 UTC DAL09 POD 6 09:00 UTC SJC01 POD 1 11:00 UTC HKG02 POD 2 13:00 UTC WDC01 POD 2 21:00 UTC WDC01 POD 3 23:00 UTC

August 26
WDC01 POD 4 01:00 UTC WDC03 POD 1 03:00 UTC WDC03 POD 2 03:00 UTC DAL05 POD 1 05:00 UTC DAL05 POD 2 07:00 UTC SJC01 POD 3 09:00 UTC SJC01 POD 2 11:00 UTC MEL01 POD 1 13:00 UTC SYD04 POD 1 15:00 UTC MON01 POD 1 21:00 UTC MON01 POD 2 23:00 UTC

August 27
MIL01 POD 1 01:00 UTC DAL12 POD 1 03:00 UTC DAL12 POD 2 05:00 UTC WDC01 POD 1 07:00 UTC DAL13 POD 2 09:00 UTC DAL13 POD 1 11:00 UTC SJC03 POD 1 13:00 UTC HKG02 POD 1 19:00 UTC

August 28
LON02 POD 2 03:00 UTC

September 04
DAL09 POD 3 05:00 UTC DAL05 POD 4 07:00 UTC DAL09 POD 4 09:00 UTC SJC03 POD 2 11:00 UTC FRA02 POD 1 21:00 UTC FRA02 POD 2 23:00 UTC

September 05
DAL09 POD 1 01:00 UTC DAL09 POD 2 03:00 UTC DAL06 POD 2 05:00 UTC DAL10 POD 3 07:00 UTC DAL10 POD 1 09:00 UTC DAL10 POD 2 11:00 UTC PAR01 POD 1 21:00 UTC AMS03 POD 1 23:00 UTC

September 06
AMS03 POD 2 01:00 UTC HOU02 POD 2 05:00 UTC WDC07 POD 1 07:00 UTC DAL05 POD 3 09:00 UTC SEO01 POD 1 11:00 UTC AMS01 POD 1 19:00 UTC LON02 POD 1 21:00 UTC TOR01 POD 1 23:00 UTC

September 07
TOR01 POD 2 01:00 UTC WDC06 POD 1 03:00 UTC WDC01 POD 5 05:00 UTC WDC01 POD 6 07:00 UTC SJC04 POD 1 09:00 UTC CHE01 POD 1 19:00 UTC LON04 POD 1 23:00 UTC

September 08
WDC04 POD 4 01:00 UTC WDC04 POD 1 03:00 UTC WDC04 POD 2 05:00 UTC WDC04 POD 3 07:00 UTC SNG01 POD 2 09:00 UTC SYD01 POD 1 11:00 UTC SNG01 POD 1 13:00 UTC

Updates

2018-08-20 13:40 UTC: Initial maintenance schedule posted

2018-08-20 15:15 UTC: Updated schedule. SYD01 POD 1, AMS01 POD 2, SYD04 POD 1, and HKG02 POD 1 updated

2018-08-22 14:45 UTC: Updated schedule. SYD01 POD 1 and SNG01 POD 1 updated

2018-08-22 15:15 UTC: Updated schedule. CHE01 POD 1 updated

2018-08-23 13:00 UTC: Updated schedule. SYD01 POD 1 updated

2018-08-24 12:50 UTC: Updated schedule. SAO01 POD 1 and SEA01 POD 1 updated

2018-09-04 16:00 UTC: Update special notice for Streaming Analytics and Geospatial Analytics services