This case study was based on our chat with Michael Buckbee, Founder of Gently Used Domains.
Introducing Gently Used Domains
Gently Used Domains helps companies trawl through masses
of data to find abandoned domains that are going to auction.
Processing Sidekiq batches for new customers
To accomplish their goals, they rely heavily on Sidekiq background job processing.
Each new signup generates on the order of 700 new unique Sidekiq jobs that need to complete before any data can be shown to the new user.
Using Judoscale to manage the load
Faced with a sky-high bill for having dozens of instances on standby (just in case someone registered that minute) or delivering a cost-effective but poor experience to new signups, Gently Used Domains turned to Judoscale.
Sidekiq worker dynos now automatically scale up as soon as a new signup is received. The entire batch is processed quickly, giving customers almost immediate access to their data.
Judoscale’s deep integration with Sidekiq queues let us easily tag which queues we wanted a faster response. We were able to tune our scaling sensitivity for exactly our usage pattern of intermittent batches of large jobs.
Realizing the cost savings
Autoscaling means that Gently Used Domains only pays for the dynos they need. They can now scale down to a single dyno when there’s little traffic, saving them loads of money on their Heroku bill.
Judoscale has been a huge strategic win for us. We’re spending less money to deliver a better experience to our users. The cost savings listed in the app dashboard is extremely conservative as the service lets us avoid tremendous backend costs on Heroku.