02-Virtualization-and-APM
Download
Report
Transcript 02-Virtualization-and-APM
P o l a r i s F i n a n c i a l Te c h n o l o g i e s
Welcomes the members of
Hyderabad chapter
for the 2nd event on 4th July 14
h e l d b y PA C E ( T h e Te s t i n g P r a c t i c e )
Virtualization
Virtualization of the
database enables compute
and storage resources to
be pooled and allocated
on demand which provides
increased flexibility, more
granular and efficient
allocation of pooled
resources, and more
scalable computing
Server
Virtualization
Server virtualization means partitioning a single server
to appear as multiple servers. Hypervisor is
responsible for creating, releasing, and managing the
resources of "guest" operating systems, or virtual
machines. Capacity planning is the art of looking at
trends in performance data to predict future capacity
needs and then building provisioning strategies around
that data.APM complements capacity planning.
Database
Virtualization
Network
Virtualization
In APE, network
virtualization enables
emulation of connections
between applications,
services, and end users
for testing
Virtualization
Application virtualization layers
replace part of the runtime
environment normally provided by the
operating system. These solutions are
widely adopted by enterprises so that
its users can securely connect to
corporate applications from any
device. But, While access
management is simplified and device
independence is enabled, IT managers
lose all ability to assess the quality of
service end users are experiencing.
(Multi-tenancy enabler)
Application
Virtualization
Storage
Virtualization
Computer Measurement Group, India
Virtualizing Storage randomizes
sequential I/O which cause a
lag in performance. APE should
develop scenarios to optimize
this performance
2
Performance Management – Evolution
APM
(Application + System)
Systems Resource Monitoring
Agent based
SNMP
Agent less
Heterogeneous
Machine Data
(Performance Metrics)
Computer Measurement Group, India
Sniffing with
Analytics
Wire Data
(Performance Metrics)
3
Monitoring Techniques – brief
1.
Simple Network Management Protocol (SNMP)
Is part of TCP/IP suite
Operates in the Application Layer – acts as an agent
Management Information Bases (MIBs) are data collectors
MIBs consists of object IDs
Each OID identifies a variable that can be read or set via SNMP
Key SNMP Protocol Data Units
GetRequest – (manager-to-agent) to retrieve the value of a variable or list of variables
SetRequest – (manager –to – agent) to change/modify the value of a variable
Trap – (agent –to – manager) to send notifications from
2.
Agent based –
System Monitoring agents run on a system to be monitored
Implementation is limited to computer systems
Agent either runs as an start-up application (or) service
RPC over TCP/IP is one of the communication medium
Agents invokes local OS commands to collect metrics
Collected metrics are saved into proprietary file formats
3.
Agent less
An intrusive monitoring system
OS commands are packaged into predefined templates
Telnet/SSH/Java RMI/WCF (xml) are used as medium to invoke commands and collect metrics
Computer Measurement Group, India
4
Monitoring Techniques – brief
4. Heterogeneous –
Combination of top-down (i.e. application ) & bottom up (i.e. system ) approach
Two sets of performance metrics are monitored
•
•
End User experience – (Active and passive) in terms of Availability and Response Times, compared to different
locations and user loads
Active monitoring (aka Synthetic Monitoring) is done by executing automated (scripted) business process flow
on an agent
Passive monitoring is usually an agent less appliance implemented using network port mirroring
Performance of Application components and system resources
Runtime Application Architecture – CMDB compliant
Application Discovery and Dependency Mapping (ADDM) solutions exist to automate the process of mapping
transactions and applications to underlying infrastructure components
Application component monitoring (Diagnostics tools)
Agent based and is generally targeted in the middleware space focusing on the application, and messaging
servers.
Provides a real time view of the J2EE and .NET stacks, tying them back to the user defined business transactions.
Shows a clear path from a code execution standpoint (e.g., springs, struts, etc.), to the URL rendered, to the user
request.
4. Wire Data
What is Wire Data – is it polling or trolling?
A scripted SNMP or Agent or Agent-less process for metrics collection is “polling”
A passive network monitoring to observe performance issues as they occur on network is “Trolling”
Polling is like “do you see performance issues”
Where as Trolling is “I see performance issues”
Computer Measurement Group, India
5
Thank you !!!
The theme and the information is sourced from the below link –
http://www.virtualizationpractice.com/going-replace-legacy-management-frameworks-23286/
Computer Measurement Group, India
6