How to deal with the recent Azure outage (Azure Websites)?

4
votes

We have TONS of websites hosted on Azure. Our VMs appear to be running now, but many of our Azure Websites are not. In an effort to bring our sites back up sooner than later, we have tried scaling UP, OUT, and changing our hosting plan, to no avail. Is there a way to force an Azure Website VM to move to another (working) datacenter? We don't want to destroy the site and bring it back up, as we will be forced to update DNS, which will cause an even longer delay in service to our customers.

Any help is greatly appreciated.

Sorry to everyone else experiencing a long night right along with me.

1
azureazure-web-app-service
According to the status updates, it looks like they're basically coming back online again... azure.microsoft.com/en-us/status – Taryn East

1 Answers

5
votes

Your best bet is to run two instances of the site in two Regions and use something like Traffic Manager (or AWS Route 53 if you want something external to Azure) to perform failover routing for you.

Depending on the type of sites you could run a static holding site in a non-Azure environment and failover to that. How you choose to solve this will depend on what your budget is (or opportunity cost in the event your sites are offline).

Note that a 99.9% yearly SLA equates to almost 9 hours of downtime in a year.

If you want to understand how you could solve this intra-Azure here's a good guide: http://blog.kloud.com.au/2014/11/03/deploy-an-ultra-high-availablity-mvc-web-app-on-microsoft-azure-part-1/