Issue Description
In a certain site, the Linux system host creates the file system on our company’s S2600 storage, and it takes over 1 hour for the process “Writing superblocks and filesystem accounting information:”.
Handling Process
1. Firstly, check if the storage has alarm information and if the initiator’s connection is normal, the outcome displays that they are all normal.
2. Check whether the host system is normal, and check whether the system version supports the file system ext4, and the outcome displays that this kind of system version also supports it.
3. The client reflects that the switch once had the phenomena of connection interruption, so we check whether the network is normal: the two switches are piled to connect with each other, one of them connects with the storage, and the other one connects with the host. Once time implemented to scan the disk, there appeared a interruption for 10 seconds, then kept on to complete the scan risk. So, we suggest the client to connect all the connections to the same switch, then create file system for each host, and they are all completed normally.
4. Via query to the researcher, we can confirm that the network interruption will cause creating the file system failed, so the problem is caused by the switches are piled and causes the network interruption.
Root Cause
Firstly, we consider the possible reasons are as followed:
1. The storage device’s problem;
2. The host layer’s problem;
3. The network layer’s problem.