sourcing operations management architecture monitoring that works!

Post on 29-Mar-2015

217 Views

Category:

Documents

1 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Sourcing

Operations Management ArchitectureMonitoring that works!

Sourcing

Operations Management ArchitectureMonitoring that works!

1. Monitoring Overview2. The Sourcing Business3. Demonstration4. The OMA Advantage5. Next Steps6. Questions

AgendaAgenda

Monitoring Overview

Why Monitor?Why Monitor?

Visibility (The Truth)

MeasurementCorrelation

Notification

Actively know environment is workingActively know when its not!Continuous visibility

Fault awareness before the end user

No finger pointing!

Half the time to resolve problemsDiagnosis is MUCH simpler

See only the problem, not the noise!

Always in touch!

Information to manage

Sweat the assets!

If you can see it, you can do something about it.Its what you cant see that will kill you!

AvailabilityIs it working?

PerformanceIs it working well?

CapacityCan it work better

without more money?

DiagnosisWhere is the problem?

Why Monitoring Systems don't workWhy Monitoring Systems don't work

– High upfront investment– Availability (and accessibility) of in-

house skills to implement and support the solution

– Technical complexity– Heavy resource impact on device and

network– Ability to rapidly deliver value– Lack of tool flexibility– Never-ending configuration,

maintenance and administration Gartner

How do we fit inHow do we fit in

OMA (Single Pane of Glass)

ESM EM EM

How we fit inHow we fit in

OMA – What we use to collect data and produce outputs

PROCESS

APPS

We will instrument and hook to applications in every way that we can

• Trusted performance and capacity advisor.

• Go to guys for performance problems

What we deliver:• Problem identification• Mission critical (apps+infr) alerting• Our Historic deliverables• Planning advice

Integrate into your operationProblem, Change, Capacity, Incident

Review, Review, Review

Weekly BaselineIncident

MonthlyBaseline ProblemIncidentAdministrationCapacity

The Sourcing Schema - Leading IndicatorsThe Sourcing Schema - Leading IndicatorsExchange What we MonitorExchange Counters Queues (6), Connectors (6), Mail (3),

Info (9), Messages (12)

Log Files (Critical Errors) (2003)

ESENT, Information Store, System Attendant, Routing Engine (10)

Operating System Disks (10), Memory (6), Services (16), CPU (5)

Hardware Temp (2), Fan (3), PSU(1), Disks(15)

Total Counters: 104 (100 Average number of counters)

• Out of a possible 10 000 counters • (Exchange has 1 700)

10 000 Counters x 100 Servers = 1 000 000100 Counters x 100 Servers = 10 000We alert on <10% of these.

99.99% of the action with < 1%of the footprint!

We poll every 2 minutes100 Tests x 100 Servers / 2 = 5 000 tests / min= 7.2 million tests / day.Imagine making it more complicated

The law of diminishing utility1st 10 tests = 50% utility2nd 10 tests = 25% utility3rd 10 tests = 12.5% utilityEtcThe 101th test = 0.01% utility= 99.99% utility.

The Sourcing Schema - Leading IndicatorsThe Sourcing Schema - Leading IndicatorsExchange What we MonitorExchange Counters Queues (6), Connectors (6), Mail (3),

Info (9), Messages (12)

Log Files (Critical Errors) (2003)

ESENT, Information Store, System Attendant, Routing Engine (10)

Operating System Disks (10), Memory (6), Services (16), CPU (5)

Hardware Temp (2), Fan (3), PSU(1), Disks(15)

Total Counters: 104 (100 Average number of counters)

• Out of a possible 10 000 counters • (Exchange has 1 700)

10 000 Counters x 100 Servers = 1 000 000100 Counters x 100 Servers = 10 000We alert on <10% of these.

99.99% of the action with < 1%of the footprint!

We poll every 2 minutes100 Tests x 100 Servers / 2 = 5 000 tests / min= 7.2 million tests / day.Imagine making it more complicated

The law of diminishing utility1st 10 tests = 50% utility2nd 10 tests = 25% utility3rd 10 tests = 12.5% utilityEtcThe 101th test = 0.01% utility= 99.99% utility.

Top 10 (50%) Next 10 (75%) Next 10System uptime Critical error logs Other servicesSystem temp Virus services % Processor timeDisk Space Other key Exchange services Disk array statusMTA services PSU & status Memory hard page faultsQueues/store Memory availableSMTP queue

Next 10 Next 10 Next 10Fan status Other memory tests Non critical logsProcessor cache status Top processors Nic errorsMail flow indicators Other CPU testsOther exchange counters

OMA

Horizontal vs. VerticalHorizontal vs. Vertical

Mom

/SC

om

BotzV

iew

10 000 tests

96 tests

We are not going to configure your Netbotz, - use the Netbotz tool.

We will tell you that your data centre is overheating, every time, only when it is. We will tell the correct person, even if you changed the configuration.

If you want 10 000 tests, get SComIf you Want 96 tests that cover the 99.99%, get OMA!

Q

How does our solution work (technical)?How does our solution work (technical)?

PESecureComms

MonitoredDevice

Are you there?How much?

A

DB

DB

DB

DB

DB

DB

DB

DB

Web

1 32

• No correlation required• No name translation required• No test differentiation required• No Contextual knowledge required• Completely “Rules” based

Sourcing Solution• Fully integrated end-to-end solution• Significantly lower resource impact on IT

environment than traditional ESM solutions

• Quick deployment ensures quick business benefit

• “Go-to-show” in less than six weeks• Instantaneous value• “Low noise” 24x7 multichannel

notification engine

Business Model• Automation as a philosophy• Source code owned and developed

internally• South African solution designed for

South African environments• Rand based pricing model• Software-as-a-Service (SaaS) delivery

model– It must work

• No Deployment, Licensing, Upgrade or Maintenance fees– Software is Free and Evergreen

The Sourcing Solution and Business ModelThe Sourcing Solution and Business Model

Our current Client portfolioOur current Client portfolio

System Inputs Technology AreasSystem Inputs Technology Areas

Operating SystemsActive DirectoryQOSNBARNAS & SANRadio LANsWan JetBizTalkDesktopsApplication ResponseMessage QueuesAssetsLog & Text FilesDatabasesDatabase Tables

System OutputsAvailability, Capacity, Performance, Diagnostics

Single Pane of GlassIncidents Technology Agnostic diagnosis

Technology Agnostic Reporting

Multi-Functional Dashboard(IT Management View)

Multi-Functional Dashboard(IT Management View)

Multi-Functional Dashboard (Executive View)Multi-Functional Dashboard (Executive View)

Function Based DashboardFunction Based Dashboard

Graphing and Trending of Performance Counters

Router DashboardRouter Dashboard

Protocol AnalysisProtocol Analysis

Application Response MonitoringApplication Response Monitoring

DNS

Network

Traffic

User Exp

User ExperienceUser Experience

Application and End User Response Monitoring

Branch ReadyBranch Ready

OMA AttendantOMA Attendant

SLA & Availability ReportingSLA & Availability Reporting

Application MonitoringApplication Monitoring

Log File Parsing

Log File Parsing

Backup MonitoringBackup Monitoring

Database Table MonitoringDatabase Table Monitoring

PESecureComms

InstrumentationInstrumentation

Counters

Counters

CountersCus

tom

App

licat

ion

Cod

e

Application Monitoring

User ExperienceUser Experience

Reporting EngineReporting Engine

IncidentsIncidents

AvailabilityAvailability

top related