Oracle focused on improving database speeds during LogFire transition
When Oracle acquired LogFire, everyone at the company knew there would be changes. LogFire managed the warehouses for hundreds of ecommerce shops, which meant juggling hundreds of PostgreSQL databases. Oracle planned at least two major shifts: integrating LogFire’s ecommerce backend tools with Oracle’s collection of web services and moving the data storage work to Oracle’s own Autonomous Database.
The legal paperwork for the acquisition was finalized in September 2016, and LogFire’s product is now known as Oracle Supply Chain Management Cloud. The database transition — moving from PosgreSQL to Oracle Autonomous Data — took a bit longer and was finally completed about 18 months ago. Oracle touts the transition as an experiment illustrating the value of its offerings. The database giant has crunched the numbers and found that its cloud supports twice as many customers at speeds 55% faster. Moreover, the automation meant the contractors who used to handle all of the manual database chores were no longer necessary.
Before the acquisition, LogFire relied on PostgreSQL databases on Rackspace hardware to track order processing for each customer. The team of contractors handled all the tasks related to patching, scaling, and failover planning and were “on call” to fix problems caused by outages. While each of these chores seems small on its own, together they required a team of 10 database administrators.
Oracle said the switch to the Autonomous Databases finished just before the pandemic hit and was essential for managing the order spikes that coincided with people staying home and depending on ecommerce shops to get the things they needed.
“They [database administrators] could add capacity whenever they wanted without downtime,” explained Oracle product management VP George Lumpkin. “They don’t need the DBA to resize their database. They don’t need to budget for or add capacity. They can add capacity on demand.”
The transition also involved moving the databases to Oracle’s own cloud with the “purpose-built” Exadata machines, which was a big factor in the faster response times Oracle reported.
“They [LogFire’s servers] moved from Postgres running on fairly generic virtual machines onto a highly optimized hardware platform with significant flash and optimized IO,” Lumpkin said. “It’s [Exadata machine] really a hardware system built for Oracle databases.”
What can other businesses expect? This largely depends upon the consistency of the database load. Ecommerce shops tend to see bigger loads during the holiday months. Some streaming services see increased demand at the end of the week. These fluctuations make work for DBAs that can be eased with automation.
For smaller companies with only a few databases, the effect of switching to Oracle Autonomous Data may not be as noticeable as it was for LogFire. The database chores normally land on some developer’s lap to be interleaved with other work.
“It’s the difference between doing everything manually,” explained Steve Zivanic, global VP of database and autonomous services. “You had to do manual tuning of the databases. You had to do manual backups. You had to essentially make sure you had enough processing power. All of that is taken care of so you can focus on the business at hand.”