Under the category "stress test," testers for Healthcare.gov ran a performance evaluation on the site. This evaluation occurred on September 30, just one day before the site was ready to hit prime-time.
During the testing, results were, quite frankly, atrocious. The tester writes (emphasis mine):
Ran performance testing overnight in IMP1B environment. Working with CGI to tune the FFM environment to be able to handle maximum load. Currently we are able to reach 1100 users before response time gets too high. CGI is making changes to configuration.
Just one day before the White House expected a flood of new users, technicians were scrambling to allow the site to handle more than a miniscule number of visitors:
To put that into perspective, the White House stated that they expected 60,000 people to visit the site on October first. In reality, 250,000 people visited. Thus, the day before it was to receive a huge influx of users, Healthcare.gov could handle precisely .44% of the traffic it got.
Even if only 60,000 people visited, as the White House predicted, the website would only be able to handle 1.8% of that number.