A system administrator would have first noticed the problem manifesting as an extremely high system load or even a system crash that could be traced back, via the normal administrative tools, to an application such as Cassandra, the Java Virtual Machine, Hadoop, or MySQL. The actual culprit, however, turned out to be a harder-to-pinpoint bug in the way Linux updated its clocks when a leap second was introduced, Ellis said.
(Curated by Dennis Moore. Read the complete article here)
Comments are closed, but trackbacks and pingbacks are open.