Heroku’s new Postgres 2.0 database service brings perks — including proactive database monitoring and alerts, and a rollback feature that is more forgiving if there’s a database snafu.
The new “unlimited undo” feature, roughly analogous to Apple’s(s aapl) Time Machine, is a big deal in the database world, said Heroku Postgres product head Craig Kerstiens. “Typically with databases, you can’t go back if something breaks. Now, if you push something to production and there’s a problem, you can now roll it back to before the issue occurred,” he said.
In addition, there are more automated and proactive e-mail alerts that give DBAs insights into what’s going on with their implementation. “If you have indexes that are not needed — they slow down your write throughput. We will now notify you about that so you can make the choice to remove them,” Kerstiens said.
Indeed, there is now a whole array or “suite” of these email alerts that do more of what a DBA would do — watch the database to make sure it’s performing optimally — but are automated, he said.
Heroku will also patch the database automatically if a critical security issue surfaces. Security vulnerabilities are a hot-button issue for virtually all software vendors, and last April the Postgres (or PostgreSQL) community had to deal with a major security vulnerability. For a managed database service to handle these patches quickly is something most users will probably appreciate, provided they’re updated on all the changes.
Heroku, the platform as a service startup that Salesforce.com bought 3 years ago, also tweaked pricing with new database tiers (see chart).