This　paper　mainly　studies　the　optimization　of　pod　scheduling　in　the　large-scale　concurrent　scenario　of　Serverless　framework　base　on　Kubernetes　platform.　In　Serverless　cloud　computing　paradigm,　rapid　deployment　and　running　of　pods　is　the　most　critical　step　to　improve　resource　efficiency.　However,　the　deployment　and　operation　of　pods　depend　on　images,　and　Kubernetes　default　scheduler　currently　base　on　pod-by-pod　mode　for　scheduling,　it　cannot　fully　meet　the　needs　of　Serverless　in　resource　scheduling.　In　view　of　this　problem,　we　propose　an　algorithm　using　the　same　pod　simultaneous　scheduling　to　further　optimize　the　efficiency　of　pod　scheduling　in　Serverless　cloud　paradigm.　Through　preliminary　verification,　we　can　greatly　reduce　the　delay　of　pod　start-up　while　ensuring　the　balance　of　node　resources.　©　2020　ACM.