Tuning Storage Foundation

  • Volker HerminghausEmail author


Having read the chapter about SAN storage, Moore's law and the advancements in disk performance your expectations about performance tuning Storage Foundation should be low. After all, optimizing a resource that is aggressively used by hundreds of different hosts for thousands of different volumes, all accessing the same overloaded mechanically limited disk hardware at the same time is almost impossible because many optimizations that improve our application performance decrease everybody else's. Fortunately, VxVM and VxFS are extensively self-tuning. It is not usually necessary to tweak individual volumes or file systems to get very good performance out of them because they know some of the features of the underlying level and use them as best they can, usually without degrading aggregate performance. For instance, VxVM knows about your storage array's sweet spots because it identifies the type of LUN using its extensive array support libraries (ASLs). VxFS in turn knows about the layout of the VxVM volume and adapts its optimization parameters (parallelity, I/O-size etc.) to the volume layout as much as possible.

However it is still possible to base one's volume layouts on wrong assumptions which might lead to very poor performance. This is the part where we can help. We therefore limit this chapter to two areas that would not be considered performance tuning in the classical sense, but that certainly classify as tuning in the sense of adapting the product to best suit your requirements as well as to prevent anything that would actually hurt performance. The main points for the Easy Sailing part are:


Queue Length File System Concat Volume Performance Tuning Short Queue 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  1. 1.Germany

Personalised recommendations