As　an　essential　concept　and　technology,　the　Industrial　Internet　of　Things　(IIoT)　is　a　feasible　way　for　the　advancement　of　industrial　digitization,　networking　and　intelligence.　However,　it　brings　serious　security　and　privacy　risks　to　industrial　data.　Blockchain　technology　can　address　security　and　privacy　risks　into　the　IIoT　effectively,　but　there　are　a　few　nonnegligible　questions　in　the　current　researches:　1)　The　scalability　of　the　conventional　blockchain　system　is　extremely　limited,　making　it　unable　to　handle　the　rapid　uploading　of　massive　data　in　the　IIoT.　2)　The　large-scale　nature　of　IIoT　and　complex　consensus　procedure　of　blockchain　result　in　excessive　computational　overhead.　3)　The　balance　between　security　and　computational　efficiency　is　often　ignored.　To　address　these　problems,　we　propose　the　concept　of　sharding　decision　blockchain　innovatively　and　formulate　the　optimization　problem　as　a　Markov　decision　process　(MDP).　Besides,　the　combining　of　mobile　edge　computing　(MEC)　and　cloud　computing　is　empowered　the　computational　efficiency　in　the　proposed　scheme.　An　asynchronous　advantage　actor-critic　(A3C)　approach　is　taken　into　account　and　used　to　solve　the　optimization　problem.　Experimental　results　indicate　that　our　proposed　scheme　provides　a　significant　improvement　on　the　IIoT　versus　other　solutions.　©　2023　IEEE.