The outage was caused by the same network update software bug that disrupted the protocol on Thursday, which resulted in nearly six hours of downtime.