VMFS blocksize considerations
VMFS alignment has an impact on performance, VMFS blocksize has no impact.
Just to clarify, there is no noticeable I/O performance difference by using a larger blocksize (i.e. 8MB).
Choosing a blocksize:
- From vSphere 4.0 onwards VMFS datastores can be expanded on the fly without the use of extents. Using 8MB blocksize for all Datastores future proofs that Datastore so it can be expanded to 2TB and allow the larger files sizes.
- VMFS blocksize should be similar within a VMware cluster, having different blocksizes will cause issues with VM’s spanning datastores those datastores. If a VM has it’s configuration file (vmx) on one Datastore, then has another VM disk on a Datastore which is bigger than the supported blocksize on the configuration file Datastore you will not be able to snapshot that VM.
Quote from the VIOPS DOC-11458
“When your Linux/Windows/Solaris guest reads/writes data, it does so in blocks and these blocks are usually around 64KB. For example, when your guest issues a SCSI read, then nothing happens to the VMFS at all – the read is “passed through” by the vmkernel straight to the SAN LUN. VMFS does not get in the way of your guest read/writes, it does not do any kind of caching, the VMFS block size is irrelevant for guest I/O.”
The blocksize does has some impact when seeking begin and end of a block / chunk. That is what the partition alignment document refers to. It’s about positioning the heads, NOT reading the blocks
Where the VMFS block size is important is when fixing the maximum VMDK size:
VMFS Block Size Maximum VMDK Size
1 Mb 256Gb