Back in the early 2010s my company was maintaining a Spring application processing ActiveMQ messages. We'd received word that the application wasn't processing work. One of us goes and looks at logs and sees just endless volumes of stack traces. Eventually the suggestion is made and accepted to just reboot the application. That fixed it.
Turns out the JVM simply lost its mind when leap seconds were introduced. So, for the next several years, we watched that French society's website that announced when leap seconds would be introduced and scheduled application restarts accordingly.
Turns out the JVM simply lost its mind when leap seconds were introduced. So, for the next several years, we watched that French society's website that announced when leap seconds would be introduced and scheduled application restarts accordingly.