(If your application is already using the High Replication datastore, you can ignore this email.)
We are scheduling a maintenance period that will impact applications running on the deprecated master/slave datastore:
Thursday 3 January 2013, 5pm US/Pacific
The master/slave datastore will be placed into a read-only period for the duration of this maintenance, though we are targeting a shorter read-only period.
Below is the information we usually send out with these notifications.
It has more information about read-only periods, how to be notified of production events and how to migrate your application to High Replication.
-Amy
----------
The deprecated master/slave datastore will be placed into read-only mode and memcache will be flushed. It is recommended that critical portions of your application that depend on the datastore should make use of the capabilities API to detect and recover gracefully during read-only periods.
Documentation describing the capabilities API can be found here:
There will be no downtime for developers whose applications use High Replication Datastore. You can read more about High Replication here:
Here is more info about the deprecation of M/S:
To migrate your master/slave application to High Replication while preserving your application ID, you can use the migration tool that is provided in the Admin Console.Read more about this tool here:
Developers can subscribe to the scheduled maintenance calendar using the following link: