Hitachi Unified Storage VM, Руководство по обзору продукта

Страница: 42 / 70

Hitachi Unified Storage VM Product Overview Guide
3–6
Software overview
primary file. A flag in the onode is used to indicate it is a cross-volume link
rather than a regular file, and an extended onode contains the information
required to access the migrated file. All of the metadata required for
directory level operations (including owner, access mode and ACLs) are
maintained on the primary file system, so operations such as "ls -I" or
"chmod" do not require access to the secondary file system. Similarly, the
information needed for quota tracking is maintained on the primary file
system, so quotas reported will include migrated files on the secondary file
system as well.
Dynamic caching
Dynamic caching reserves space on a storage tier for caching of "hot" files
accessed through NFS. The space reserved is actually an entire file system
unto itself, and as such can be as large as any other file system in the
namespace. By definition, any file that is recently accessed may have a copy
also located in the dynamic cache. If the cache is created in a high-
performance tier of storage, this copy guarantees that any hot files are
automatically on the highest performance disk tier, which may actually be
an SSD or a hybrid SDRAM/SSD tier. Having the cache prevents the need
for reverse data migration.
Cluster read caching
Cluster read caching is dynamic caching applied to a single server or to a
cluster of servers. If applied to a single server, the feature is called local
read caching. When used with a cluster of servers, each server maintains
its own dynamic cache, but is aware of the files accessed by all the other
servers in the cluster. Copies of hot files from anywhere in the cluster
therefore make their way to every cache on every server, which can result
in dramatic aggregate read performance improvements. This is because
every server can respond to any read request for a given set of hot files. In
this way, dynamic caching works with the data migrator to simultaneously
provide policy-based data movement in both the forward and reverse
senses.
Data relocation
The file module system allows you to relocate date in three ways:
•
EVS migration is typically used for load balancing among servers or
cluster nodes, and vacating a server for scheduled maintenance. It
makes it possible to relocate a virtual server within a cluster, or to a
server outside of the cluster that shares access to the same storage
devices. EVS migration has minimal impact on network hosts and, once
it has completed, those hosts may access the data using the same path
names that were in use prior to the relocation.
• File system relocation : Any file system accessed via cluster
namespace can be relocated to another server within the cluster. File
system relocation has minimal impact on network hosts; once
completed the data is accessed using the same path names that were in
use prior to the relocation. File system relocation is typically used to load
balance within the unified namespace.