Upload
johnnie-burke-gaffney
View
110
Download
4
Embed Size (px)
Citation preview
Managing the Virtual NightmareFor Dynamic Cloud, Virtual and Physical Infrastructures
Presentation by Johnnie Burke-Gaffney
& Stuart Kennedy From eG Innovations
© eG Innovations, Inc | www.eginnovations.com
About eG Innovations
© eG Innovations, Inc | www.eginnovations.com
About eG InnovationseG Innovations is a leading provider of enterprise-class performance management solutions that provide complete visibility across every layer and every tier of dynamic & complex cloud, virtual, and physical IT environments to reliably deliver mission-critical business services.
Locations USA, UK, Netherlands, Singapore, India
Customers Over 1000 customers worldwide
Employees 350
Year Founded 2001
Certifications VMware Ready, Citrix Ready, SAP Certified, Red Hat Certified
© eG Innovations, Inc | www.eginnovations.com
Are You Experiencing The Nightmare?•High cost of downtime•Slow applications•Poor user experience
User Frustration & Low Productivity
•Manual, complex and lengthy diagnosis•Requires many domain experts
RisingIT Support
Cost & Complexity
•Throwing hardware at problems•Oversized & inefficient environment
RisingInfrastructure Cost
•Delayed deployment & rollout•Performance issues & fire fighting•Failed projects & initiatives
Delays, cost overruns, missed
objectives
Comm
on Symptom
s ...
© eG Innovations, Inc | www.eginnovations.com
VirtualizationPerformance Assurance
© eG Innovations, Inc | www.eginnovations.com
Bill payment from my internet banking
account isn’t working.
The CRM service is slow!
My online flight reservation
did not go through.
Users care about “services”
The CPU usage ofthe Linux servers
is ok.
The DNS servers areresponding well to
queries.
IT operations teams focus on infrastructure silos
This disconnect is a threat to the success of transformational IT initiatives & the promise of agility, scalability, and cost savings!
The User / IT Management Disconnect
The User Experience Challenge
© eG Innovations, Inc | www.eginnovations.com
The “It’s Not Me …!” Syndrome
End User
Client Admin
LAN Admin
Firewall admin
Server admin
Virtualization admin
Domain admin
ERP Admin Sys admin ApplicationAdmin
The serveris working
OK
No othercomplaints
All lights Are green
We don’t see anything
wrong
Database Admin
Hey, this is not working
VMs are lightly loaded
EverythingIs OK
Not ourproblem
Looks fine Not mine
eitherTalk to
the otherguys
IT Service Manager
© eG Innovations, Inc | www.eginnovations.com
FIREWALL WEB SERVER
USER
Suppose the database server is 50% slower than normal
APP SERVER DB SERVER
Login
Register
Brow
se
End to End Service - Cause and Effect
A problem in one tier can affect all the other tiers involved in service delivery
© eG Innovations, Inc | www.eginnovations.com
Disk reads
Streaming Media AppSlow Database Queries
Virtualization Breaks Management Ground Rules
Excessive disk reads by the media server slow down Oracle database accesses
Virtual infrastructures are hard to manage. Traditional monitoring tools are not designed to handle these dynamic environments.
© eG Innovations, Inc | www.eginnovations.com
Where Time & Money is Being Spent
© eG Innovations, Inc | www.eginnovations.com
eG Enterprise Service Manager
Identify & resolve problems preemptively,
before users call!
Network? Database?
Application?
VMware?Storage?Profile Server?
The Service Manager is a General Practitioner for your IT infra.
eG Enterprise ManagerBusiness Service Owner
© eG Innovations, Inc | www.eginnovations.com
Pinpointing the Root-Cause Diagnosis for Virtual Application Slowness:
A Real-World Example of How eG Enterprise Helps
© eG Innovations, Inc | www.eginnovations.com
Real User Transaction Monitoring
Response time metrics for the web-based service: ISG_WEB
Checkout and TransferBalances transactions have unusually high response times
Clicking on any of these transactions displays the service topology diagram for this web-based service
© eG Innovations, Inc | www.eginnovations.com
End to End Root-Cause Diagnosis
• Know which tier of a business service is impacted
The dependency arrows and color coding make it clear that a problem with the MS SQL Server is impacting the web server.
© eG Innovations, Inc | www.eginnovations.com
Virtualization-Aware Root-Cause Diagnosis• Know where the root-cause of a problem lies:
The SQL Server VM is hosted on an ESX Server, and something in the ESX Server itself is impacting the SQL Server VM.
Clicking on this icon brings up the layer model for the ESX server.
© eG Innovations, Inc | www.eginnovations.com
Best Practice Virtualization Monitoring
Something is wrong with CPU usage of the ESX console.
The ESX console is taking up close to 50% of the server’s physical CPU, which is very unusual !
• Know which layer is impacted – Network? System? Application?
The problem is at the OS layer.
Clicking on the diagnosis button lets us find out why.
© eG Innovations, Inc | www.eginnovations.com
Virtualization-Aware Root-Cause DiagnosisList of the top 10 CPU processes running on the vSphere/ESX service console
A Samba backup job is using almost 95% of the ESX console’s virtual CPU !
This is the root-cause of the web response time issues !
© eG Innovations, Inc | www.eginnovations.com
eG Patented Root-Cause DiagnosisWithout root-cause diagnosis, you have no idea where the problem lies
The root-cause of the problem
The effects of the problem
Simply clicking on this diagnosis button shows the root-cause of the problem: the Samba issue shown in the previous slide
All the problems appear to be equally important.
With root-cause diagnosis, you have a clear idea of what to do to resolve the problem.
© eG Innovations, Inc | www.eginnovations.com
The ROI of Performance Assurance•Reduce downtime•Increase application availability•Boost user experience
Boost User Satisfaction
& Productivity
•Simplify, automate & accelerate diagnosis & troubleshooting •Optimize staffing levels & reduce OPEX
Reduce IT Support
Cost & Complexity
•Increase hardware utilization •Leverage investment in software•Right-size & optimize environment
Reduce Infrastructure Cost & Avoid Cost Overruns
•Accelerate deployments & rollout•Reduce risk, deliver successful projects & peace of mind
Deliveron Time, on Budget,
on Target
Comm
on Results ...
© eG Innovations, Inc | www.eginnovations.com
eG Enterprise
© eG Innovations, Inc | www.eginnovations.com
eG’s Key Technologies
© eG Innovations, Inc | www.eginnovations.com
The eG Universal Agent
• A single agent license for Microsoft, Linux, Sun Solaris, HPUX,IBM AIX, VMware, Tru64
• A single price, regardless of OS or server configuration - 2, 4, 8, 16 CPUs
• A single agent for monitoring any application
• A single price to manage multiple applications on the same server
• Auto-upgradeable• Agentless monitoring option• 100% web-based – HTTP/HTTPS
© eG Innovations, Inc | www.eginnovations.com
Monitoring Every Layer/Every TierComponent Type Applications Monitored by the eG Suite
Web Servers Apache, iPlanet/SunONE, Microsoft IIS, IBM HTTP Server , Oracle Http
Web Application Servers WebLogic, ColdFusion, ATG, iPlanet, SunONE, Microsoft transaction server, WebSphere, SilverStream, JRun, Orion, Tomcat, Oracle 9i OC4J, Borland Enterprise
Enterprise Applications SAP R/3, SAP ITS, Corillian Voyager, Micros Opera, Oracle Forms, SiteMinder
Database Servers Oracle, Microsoft SQL server, DB2 UDB, Sybase, MySQL, Informix
Terminal Servers Microsoft Terminal Server, Citrix XenApp
Network Devices Cisco routers, Cisco Catalyst switches, Baystack hub, Network nodes, Local Director, Cisco VPN Concentrator
Microsoft Applications Active Directory, BizTalk server, Windows Internet Name Service (WINS), DHCP server, MS Print server , MS Proxy server, MS File server, ISA Proxy server
Firewalls Check Point Firewall –1, Cisco PIX, Juniper Netscreen
Email Servers Microsoft Exchange, Sun ONE messaging, Lotus Domino, Qmail, Sendmail
Messaging Servers MSMQ, IBM MQ, FioranoMQ server
Others FTP, MTS, Event Logs, Tuxedo domain servers, Printers, NetApp Filers and NetCache, SiteMinder Policy server, Radius server, COM+ server, ASP .NET server,
Operating Systems Windows NT, 2000, 2003, 2008, 2012, 7, XP, Solaris, Linux, AIX, HPUX, Netware, OS400
Virtualization Platforms VMware vSphere, Citrix Xen Server , Solaris Zones/LDOMs , Microsoft Virtual Server
VDI Connection Brokers Citrix XenDesktop, VMware View, Leostream CB
© eG Innovations, Inc | www.eginnovations.com
The eG Virtualization Monitor• The Outside view shows the portion of physical
resources used by each VM (CPU, disk, memory)
• Provided by the virtualization hypervisor
• Useful for capacity planning and identifying certain VM issues
• Does NOT show why a VM is consuming resources
Reso
urce
s of
the
Phys
ical M
achi
ne
0
100%
VM1 15%
VM2 25%
VM3 20%
VM432%
0
100%
Reso
urce
s of
the
Phys
ical M
achi
ne
VM1 15%60% 10%
VM2 25%10%
45%
5%30%
VM3 20%25% 60%
VM4 32%12%20%
40%
Apps inside a VM
• The Inside view shows the portion of resources allocated to a VM that are used by each application and each user of the VM
• Provided by the guest OS (for Windows: WMI)
• Useful for user load balancing, identifying guest OS issues, misbehaving applications, and unauthorized user activities
• Does show why a VM is consuming resources, accelerates fix
© eG Innovations, Inc | www.eginnovations.com
Extending eG For Monitoring Custom Applications
© eG Innovations, Inc | www.eginnovations.com
Auto-Baselining of MetricsMost operators have too much data. They need “information.”
Automatic time-varying baselines – make configuration simple, and monitoring PROACTIVE
© eG Innovations, Inc | www.eginnovations.com
Integrating Performance & Config Management
PERFORMANCE ALERTS
Track configuration changesCorrelate performance with configuration changes
CONFIGURATION CHANGE
Benefit: Saves endless hours of troubleshooting
© eG Innovations, Inc | www.eginnovations.com
eG Value Proposition
Proactively detect and correct problems before users notice Increase revenues by reducing mean time to repair Efficient use of operations staff
With
eG
Problem
Resolved
Problem Occurs
Problem
Isolated
Large amount of time saved
Problem
Resolved
Problem Occurs
User Notic
es
Slowdown
80% of time spent in isolating the problem
TOD
AY Problem
Isolated
Mean time to Repair (MTTR)is very high
eG Enterprise
Proactively detect and correct problems before users notice Increase revenues by reducing mean time to repair Efficient use of operations staff
© eG Innovations, Inc | www.eginnovations.com
ROI Example Without
eG InnovationsOverallImpact
Reduce Downtime per Occurrence by 90%
180 minutes x £5,000
= £900,000
20 minutes x £5,000
= £100,000
~ 90% savings per outage(£800,000)
Reduce Outage Frequency & Cost by 91% (annual)
20 outagesx £900,000
= £18,000,000
16 outages x £100,000
= £1,600,000
91% + savings per year
(£16,400,000)
Reduce IT Support Cost by 15% (annual)
20 FTE x £80,000
= £1,600,000
17 FTEx £80,000
= £1,360,000
15% savings per year
(£240,000)
Improve User Density on HW by 20%
100 users / server e.g. 500 servers
120 users / server e.g. 300 servers
20% HW server savings
Accelerate Time to Deployment by 20%
100 Hours(1,000 desktops)
80 Hours(1,000 desktops) 20% faster
Boost User Experience More productive
© eG Innovations, Inc | www.eginnovations.com
eG Performance Assurance Benefits
Accelerate adoption rates
Enhance service uptime
Achieve great ROI
Deliver great user experience
© eG Innovations, Inc | www.eginnovations.com
ROI – How? Product Features
Reduce Downtime per Occurrence by 90%
• Earlier alerting• Faster diagnosis due to better
visibility and auto-correlation
• Broad and deep cross-domain visibility• Auto-correlation & rapid , precise
diagnosis from user to root cause
Reduce Outage Frequency & Cost by 91% (annual)
• Pre-emptive alerts before users are impacted
• Rapid diagnosis and fix
• Intelligent baselining• Pre-emptive alerts• Actionable diagnostic intelligence
dashboards
Reduce IT Support Cost by 15% (annual)
• Fewer calls to helpdesk• Fewer incidents to troubleshoot• Easier to troubleshoot / fewer
domain experts
• Actionable alerts & auto-diagnosis dashboards & reports
Improve User Density on HW by 20%
• Deeper visibility into resource utilization and user impact (both over-capacity and bottlenecks)
• Add more users to existing infrastructure
• Capacity and trending reports
Accelerate Time to Deployment by 20%
• Identify bottlenecks early• Avoid performance issues• Deliver on time, on budget
• Actionable alerts & auto-diagnosis dashboards & reports
• Capacity and trending reports
Boost User Experience• Proactively monitor user experience • Proactive alerting & diagnosis• Get more productive users
• Pre-emptive alerts• Auto-correlation
& diagnosis from user to root cause
© eG Innovations, Inc | www.eginnovations.com
For More Information:http://www.eginnovations.com