Block level VS File level storageadmin | Tuesday, December 8th, 2009 | 2 Comments »
I wanted to provide a quick summary of the differences between block and file level storage and how it pertains to VMware. This is not going to be a NAS vs SAN article, but more of the fundamental differences between how they each store data.
Block level data storage
- Protocols that use block level: Fibre channel, iSCSI, FCoE etc…
- Does not have a file system such as ext3 as overhead.
- SCSI commands are sent between the initiator and target.
- Data block sizes are fixed.
- Storage file system can be VMFS.
- Generally a more expensive storage type versus NAS devices that typically use file level data storage.
- Generally faster than file level storage. (although things are getting closer)
File level data storage
- Generally devices that use file level are NAS, and typically cheaper than SAN’s that use block level data storage.
- Uses the file system ext3.
- Data is written and read as variable length files. These files are broken down into blocks and then stored onto disk. (this is your overhead)
- Cannot support VMFS format.
- Cannot support external bootup for ESX/ESXi host servers.
- File transfer protocols: NFS(v3), CIFS, HTTP/DAV, FTP, rsync, etc…
- Generally slower than block level storage. (although things are getting faster & better everday)
So there you have it. A quick run down on block level VS. file level. I’ll continue to update this article as I run into more useful data points.