File System Block Size while creating the File System using mkfs

342 views Asked by At

I am trying to use BUSE (with NBD) to create a block device in user space. I am not clearly understanding the block access patterns when creating a file system. As shown in the example when I mount the nbd device and create a ext4 file system with a block size of 4096, I am seeing the reads and writes are in multiples of 1024 and not 4096.

However once the file system is created, when I mount the device and try to read/write files the requests are being sent in multiples of 4096.

So it looks like, while creating the file system using mkfs.ext4, the block device is accessed with 1024 as block size and only after the file system is created, the user specified block size will be used. I am correct in making this inference? If so, can someone explain what happens at the backend and why 1024 is chose initially?

Thanks and Regards,

Sharath

0

There are 0 answers