5.2

16 vFabric Suite
16 Overview of vFabric Suite 5.2
database buffer pool to make memory available, and, when memory is available, accordingly increases the amount of memory
dedicated to the buffer pool.
• Higher default values for many critical settings compared with standard Postgres databases, including the settings for
shared_buffers, checkpoint_segments, and wal_buffers. These defaults give you higher out-of-the-box performance than
standard Postgres and a database that you can tune more easily according to workload. (A slight tradeoff occurs in disk space
and memory usage.)
• Checksums performed by default on each write operation to tables and indexes, ensuring that when vPostgres retrieves data,
the data is clean.
• Allows you to configure the required crash-recovery Service Level Agreement (SLA), and tunes the checkpoint parameters
dynamically as it monitors performance. This feature addresses the trade-off between configuring more frequent checkpoints
(for less crash recovery time and lower performance) and configuring less frequent checkpoints (for higher performance and
more crash recovery time). Solving the trade-off problem in the virtualized world is especially complex because resources
alloted to the database virtual machine vary dynamically.
For new feature information, see vFabric Postgres Release Notes.
vFabric Application Performance Manager 1.1
Designed specifically for application deployment in hybrid clouds and virtual environments, VMware® vFabric™ Application
Performance Manager (APM) integrates VMware® vFabric™ AppInsight and vFabric Hyperic to improve service levels of
applications (availability and performance) and reduce cost. An APM license includes license keys for vFabric AppInsight and
vFabric Hyperic. You can also register adapters with vCenter that enable integration of AppInsight with vCenter Orchestrator and
vCenter.
Main features of vFabric Application Performance Manager:
• Single unified interface that provides at-a-glance health state for an application. From there you can easily zoom in on
problematic areas and apply one or more remedial actions to all code levels, middleware, and infrastructure.
• Automatic identification of business-critical transactions.
• Ability to track real-time changes in transaction patterns and inventory, including the starting, stopping, and moving of virtual
machines.
• Out-of-the-box baseline to enable the identification of normal as compared to abnormal behavior.
• Visualizations and drill-down details, such as correlated metrics on resource performance from vFabric Hyperic and byte code
instrumentation, make it easy to identify root cause and take actions at the code level to fix problems quickly.
• Ability to view the correlation between application infrastructure changes and performance changes.
• Can trace the complete flow of a transaction remotely cross all layers of the application, regardless of where it is deployed.
• Computed health score derived from key indicators such as throughput, latency, hit rate and error rates to provide early
warning for building performance problems.
• Integration with vCenter Orchestrator provides out-of-the-box workflows that help administrators automate existing manual
tasks. vCenter Chargeback integration improves utilization of virtual infrastructure with accurate visibility into the true costs of
virtualized workloads.
For new feature information, see vFabric AppInsight 1.1 Release Notes and vFabric Hyperic Release Notes.