Don’t most people agree that it tends to get confusing when considering block\chunk size and alignment? I would say most people do because there are many factors to take into account when building a solid storage design.
What is affected in the near term?
· SAN LUN
Mostly, I wonder why particular vendors seem to be more affected by misalignment than the others. I just consider it to be directly related to the cluster\block sizes on the storage back end.
The following is an excerpt from the Microsoft VHD Performance White Paper:
“Certain features existing in SAN environment may impact actual VHD performance results. For example, you may see performance penalties caused by misalignment of dynamic VHDs on certain type of SANs which may not show up on a DAS platform”
The following Executive Summary from VMWare states that it is necessary to align partitions on both physical machines and VMware VMFS partitions to prevent performance I/O degradation:
What does the Executive Summary mean above in more detail?
· EMC Clariion it’s a different story. Clariion arrays use a 64KB chunk size to write their data which means that not every Guest OS cluster\block is misaligned and thus EMC Clariion is less effected by misalignment. The suggestion they mention is to align directly with new Server OS’s using a 1024 offset, although it isn’t necessary.
· NetApp currently uses a 4KB block size for the hardware. So, this would mean that if the file system uses a 4KB cluster\block size as well. With this, it means every single IO will require the array to read or write to two blocks instead of 1 when your VMDK’s are misaligned.
This is a link referencing the new 4KB hardware block size and how it relates to Windows Advanced format (4K) disk compatibility update (Windows)
Here’s some alignment graphics from VMware with the article referenced