Cloud Bigtable IAM and Stackdriver metrics-based scaling

[Misha Brukman]

[ bcc: google-cloud-bigtable-announce ]

Dear Cloud Bigtable users,

I'd like to give you an update of some recently-released features and functionality for Cloud Bigtable.

First, we've recently rolled out IAM roles for Cloud Bigtable, so now you can provide access to just the Bigtable instances in a project to a developer or a service account, without also giving them access to all other resources (as you would if you used project-wide permissions owner/editor/reader).

We are also exporting Cloud Bigtable server-side metrics to Stackdriver, which you access via Stackdriver Metrics Explorer and also programmatically access these metrics. As per the note on the page, please be advised that Stackdriver Monitoring alerts and custom dashboards are not currently available for Cloud Bigtable usage metrics; this support is in the works and will be available in the near future. Note that you can also measure client-side metrics if you are using the Java HBase client.

We've heard from several Bigtable users that they would like to use the server-side metrics such as CPU utilization to build custom autoscalers, so we developed sample code for programmatically scaling Bigtable with Python and Java.

Finally, we've published a new page with articles, videos and blog posts featuring Bigtable architecture and use cases, and we've added several more integrations as well.

We'd love to hear your feedback — please let us know your thoughts on these features and what you'd like to see next!

Misha