News

Virtual network maintenance in regions(SG1,UK1,US1,DE1,HK1) 


Thank you for using NTT Communications Enterprise Cloud  (ECL) service.

We will conduct maintenance for virtual network service in region (SG1,UK1,US1,DE1,HK1).

This maintenance is to improve service quality especially for network infrastructure of Enterprise Cloud 2.0(ECL2.0).

Thank you for your understanding.

 

The details are as follows:

Target region:

Region:

SG1,UK1,US1,DE1,HK1

Maintenance Schedule:

SG1(Singapore)region:

Mar./22/2019 (Fri.) 12:00 UTC ~ Mar./24/2019 (Sun.) 22:00 UTC (Fall back time included)

 

UK1(United Kingdom Summer time)region:

【Plan】Apr./19/2019 (Fri.) 19:00 UTC ~ Apr./22/2019 (Mon.) 5:00UTC (Fall back time included)

 

US1(USA)region:

【Plan】May./17/2019 (Sat.) 00:00 ~ May./20/2019 (Mon.)10:00 UTC (Fall back time included)

 

DE1(Germany)region:

【Plan】Jun./14/2019 (Fri.) 18:00~ Jun./17/2019 (Mon.)4:00 UTC (Fall back time included)

 

HK1(Hong Kong)region:

【Plan】Jul./5/2019 (Fri.)12:00 ~ Jul./7/2019 (Sun.) 22:00 UTC (Fall back time included)

 

Affected Services during maintenance window

1.GUI(ECL2.0 portal)and APIs are not available*

In this maintenance window, GUI/API configuration to the resources of regions which is in maintenance. In the term of maintenance window, the access is limited only via console connectivity to Virtual Servers and Baremetal Servers. We highly recommend not to connect or conduct resources just before the maintenance window, as there might be affection to use services. On the other hand, it is available to connect to other regions via GUI/API.

 

2.Interruption of BUM traffic:Less than 1 min.

One time per VM* (in case of failback as well)

Affected services :Virtual Server, Firewall, Load Balancer, Security (Managed)

And Services connected above. e.g., Block Storage connected Virtual Server Broadcast, Unknown unicast and Multicast may be unstable during the maintenance.However, there may be no impact for services as we have functionality of retransmission the traffic.

Regarding the redundant configuration, FW and LB may switch to the redundant side.Any query regarding the case, please refer to customer support via ticket.

 

Following is detailed duration which will be affected to unicast per region

1.SG1

・Group A

Start time: Mar 23th(Sat) 1:00– 5:00 UTC*

・Group B (will be started 2 hours after completion of the work at Group A)

Start time: Mar 23th(Sat) time between 4:00 – 8:00 UTC*

End time: Mar 23th(Sat) time between 6:00 – 10:00 UTC*

UK1,US1,DE1,HK1   To Be Determined

 

FAQ

Q:Are there any operations required from customer side before or after the maintenance?

A: Switch between active and redundant side of FW and LB may occur during the maintenance. We recommend customers to confirm the redundant configuration (VRRP related and DHCP settings in logical network) before the operation.

 

Service details are in the following links:

Managed Firewall/UTM

https://ecl.ntt.com/en/documents/tutorials/security/rsts/security/operation/managed_firewall_utm/3120_interface_ha.html

https://ecl.ntt.com/en/documents/tutorials/security/rsts/security/operation/managed_firewall_utm/8080_ng_setting_guide.html

 

Load balancer

https://ecl.ntt.com/en/documents/tutorials/rsts/LoadBalancer/network/lb_vrrp.html

 

Firewall

https://ecl.ntt.com/en/documents/tutorials/rsts/Firewall/vyatta/command.html

 

Known issues

https://ecl.ntt.com/en/known-issues/vrrp-configuration/

 

②Configuration to connect between virtual server and block storage

As for traffic from virtual server to block storage, it occurs interruption of unicast at virtual servers but in case that official template and users set configuration following our tutorial, disk I/O communication is not to be interrupted by multi-path device functionality, even when iSCSI connection is to be interrupted.

However depending on application there might be some interruption hence please check the configuration at each tenant by customer.

Also we highly recommend that customer who don’t set multi-path configuration, change the configuration of time-out amount of iSCSI initiator as “120 seconds”. (Please check impact to each applications in case of change configurations.)

On the other hand, there’s no impact to connectivity between Virtual Server instance and Volume, and Baremetal Server and Block Storage as well.

 

 

Q:Is it possible to operate VMs during maintenance?

A: During maintenance, it is not available to create/refer/update/delete VM via portal or API.

SSH connection to VM may be influenced due to any network service impact caused by the maintenance. We recommend customer to operate VM out of maintenance window.

 

Q:please describe the impact on the application layer due to the network service impact caused by the maintenance.

A: We apologize that we cannot provide detailed information since the impact in application layer are depending on each customer’s environment.

 

Q:Which service menus are affected due to the interruption of multicast at this maintenance duration?

A: There is interruption for logical network which connect to Virtual Server, Firewall, Load Balancer, Security (Managed). Hence in case of the source/destination of the traffic are from/to these service components there may be affected to use, such as Baremetal Server, Storage, Internet Connectivity, VPN Gateway, SD-Exchange, Colocation Inter-Connectivity (CIC), Enterprise Cloud 1.0 Inter-Connectivity (EIC), Amazon Web Services Inter-Connectivity, Google Cloud Platform Inter-Connectivity, Microsoft Azure Inter-connectivity, Enterprise Cloud 2.0 Inter-Connectivity, DC Inter-Connectivity).