As @richgalloway already said this is known issue. It could be also false positive as tresholds are too low.
There are several articles and discussions of it already. I link just some here:
The threshold for IOWait alerts defaults to too low a value. Work with your system admins to determine the best value for your system.
If this is a physical machine - check your actual I/O statistics and - if you're really overstressing your hardware - get better one. If this is a VM, check with your virtualization and storage teams if you're overstressing your storage but most probably you'll just want to disable that health check - with VMs it most often just lies.