one of my production server report to me the sudden speed downgrade I checked the mongod log and found this at that moment

Mon Jun 6 20:41:32 [FileAllocator] done allocating datafile /data/dt_master_0/dt_base.7, size: 2047MB, took 372.415 secs

It seems it took about 6 minutes to accept the next query, is there any way to improve this, or I can manually allocate the disk for them?

I just can't afford the sudden strike during the production time