Google App Engine, the search giant’s cloud application development and hosting platform, experienced a Java outage late last week that resulted in high latency and error rates. By Google’s reckoning, only 1.9 percent of total App Engine traffic was affected. But it’s still a blow to Google App Engine’s credibility as the company tries to build cachet with enterprise ISVs.

The outage began at 7 p.m. Pacific on July 14, 2011, and gradually increased in magnitude until Google engineers were able to begin repairs at 9:30 p.m. The Google App Engine blog explaining the outage says that the repairs were completed by 11:30 p.m., and applications were back to normal operation. The outage occurred very shortly after a scheduled maintenance period, but Google insists that the two events are unrelated.

What I found most interesting about Google’s blog entry was the apologetic tone. See for yourself:

When Amazon EC2 infamously went down for a few days earlier in 2011, Amazon Web Services wasn’t nearly half as apologetic as Google seems to be about some four-hour, comparatively low-impact downtime. Regardless, it’s not good PR for an offering that Google is pitching as ready for primetime.