Hey there! This is a case about exceptions that in 12 hours about 9TB space is used.
【Problem Description】：
We had observed that 12 to 13 TB are available in the morning, but after about 12 hours, they had found that only 3 TB storage is available. Even though we didn't make many provisions today but observed that storage had fluctuated a lot.
【Problem Analysis】：
The change amount of data is related to 3 operations: write/delete/modify.
In 20190709, from 9:30 am to 21:30, the average write bandwidth is 107.02MB/s in controller A, and in controller B the write bandwidth is 116.81MB/s.
It means, in 12 hours there are 9.22TB written into the storage. 4.409088134765625 TB(=107.02MB/s * 3600 s *12 hours) + 4.812423706054688 TB (=116.81MB/s * 3600 s *12 hours )= 9.221511840820313TB.
Because there are snapshots, although the upper layer hosts delete/modify some data, the old data will be still in the storage space.
That is why “We had observed that 12 to 13 TB are available today morning & now we had found that only 3 TB storage is available”.
Controller A:
【Solution Description】：
We wrote much data into storage, it's not the Huawei product problem. We do not know how much data the end users write. So the only way is to check the storage performance data.
Hope can help you!
