Scaling with Cloud Spanner

[@googlecloudtech]

Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech

[@akashsrivastava9716]

Great content!!! Kudos to the team

[@clivemthethwa3576]

Hi Im assuming changing from processing units to nodes or vice versa is not an issue, am i right?

[@clivemthethwa3576]

Will it not cause disruption?

[@justinmakeig5170]

A node is just 1,000 processing units. They’re different units measuring the same thing. Below a full node (1,000 processing units), you can scale by increments of 100 processing units, all the way down to 100, the equivalent of 0.1 nodes. Above 1,000 PUs, or a full node, you can scale by increments of whole nodes. Changing the number of processing units in an instance is always an online operation, regardless of how many you’re using.

[@sandeepkushwaha5716]

9:18 Does cloud spanner supports NoSql dbs?

[@mohitgulati4614]

Many scaled workloads which today run on NoSQL DBs but need relational semantics are a good fit for Spanner. Many of Spanner's customers have migrated from DynamoDb, Cassandra, MongoDB and CosmosDB.