A government customer purchased two Huawei all-flash storages for business deployment. During the use of the storage, engineers found that the storage performance often fluctuated. If the engineer…

A government customer purchased two Huawei all-flash storages for business deployment. During the use of the storage, engineers found that the storage performance often fluctuated. If the engineer wanted to locate the storage performance fluctuation problem at this time, it would involve many operations. 1- Check whether there are obvious changes in IO model or business load at the fluctuation time point 2- Analyze the latency fluctuation module through the latency location tool FTDS on the storage side 3- Check whether there is an obvious hit rate mutation at the fluctuation time point 4- Check whether there is a very small IO scenario at the fluctuation time point 5- Check whether there is any obvious hardware resource bottleneck on the storage side at the fluctuation time point 6- Check whether there are obvious storage background tasks at the fluctuation time point I would like to ask, among the above operations, which operations can locate the problem and which operations cannot?