Compartmentalized Common Operational Picture Sc2 Op
-
Upload
httpsawconceptscom -
Category
Documents
-
view
263 -
download
0
description
Transcript of Compartmentalized Common Operational Picture Sc2 Op
Sync’d Compartmentalized Common Operational Picture: SC2OP
1
SAW Concepts LLC Modified Presentation to NYC Computer
Measurement Society, Summer 2011
S.O.S
SYNCSEED_CLOUD
Situational Awareness CDRLS
.0001 .05 .01 .1 5 15 30 99 .0001 .05 .01 .1 5 15 30 99
SYNCSEED_CLOUD
The “SC2OP”
Avada Confidential - Feedback Monitoring - 2010-11
2
A complete, integrated, secure, collaborative portal which
facilitates logical, virtual environments to aid with problem
determination, administration, monitoring, testing, auditing
& statistical reporting for Message Oriented Middleware
(WebSphere MQ, Tibco, & any JMS base solution)
* Providing collaboration between business units
* Providing collaboration btwn different MOM Environments
Avada Software Markets Infrared360
SC2OPSynchronized, filtered
Single Integrated Compartmentalized
Operational Picture derived from the Family of
Interoperable Operational Pictures (FIOP).
Role based / filteredCOMPARTMENTALIZED
Common Operational Picture
Avada Confidential - Feedback Monitoring - 2010-11
3
Metrics For Business Units
Feedback Monitoring – Working with Business Units to Determine Business Application SLA Criteria
Topics:1.Dangerous assumptions planning THRESHOLDS for new business applications2.Best practices for planning threshold management for new business applications
Feedback Monitoring:-Heartbeat msg update: y/n-Update = new netconfig
planning thresholds for new business applications> / < threshold by intensity, duration, magnitude… threshold rules dictate alert and event feedback to mitigate / adjust & reform orgs..
DECISION
CAPXML
Avada Confidential - Feedback Monitoring - 2010-11
4
Business Scenario
Airline plans promotion via new business partner.
New business partner is actually subsidiary and previously booked through airline, but not by leveraging their real-time systems.
Business unit announces new partnership, but not details of new business application.
Business unit understands that middleware group has handled this type of data transaction before.*
Business unit and middleware group do not meet to review new application criteria, only details of middleware objects needed.
Business unit assumes similar thresholds will monitor this application the same way; “this one’s not too different!”
DoD systems: driven by Structured military message Threads that along with theIndividual FFURNS and FFUDS[data elements as 3 / 4 digit codes] Determine funding and developmentSchedule and pace – often, single data elements take over a year to NOT change – e.g., the infamous redThread where an AVIATION data element was the fly in the ointment between a ground system and a intelligence system – true story!
SoS
SABRE
Avada Confidential - Feedback Monitoring - 2010-11 5
Example Application Environment
Multiple Queue Managers with multiple Queues associated with “each” different business application.
Different business applications exist on the same Queue Managers - in order to leverage server capabilities.
Different applications built by different business units, different development staff, different business requirements, different system footprint - all on same server!
Administrators assume similar thresholds will monitor all of the different applications the same way; “this one’s not too different!”
The Heart Beacon involves pre and post configuration management of the network and is neutral to systems and middleware. What Vinton Cerf and his team described metaphorically as the heartbeat is actually a interval in time where (state meta) data is gathered and placed into temporary holding areas, queues, file structures for onward delivery by middleware software such as agents, bots, motes, scripts. The heartbeat is an opportunity in time to synchronize data harvesting protecting bandwidth on low capacity networks prior to entry into portals and fusion centers – the point of no return in context with event time stamping. After transition to the middleware or cloud layer, it is too late to try and re-establish timing of event / alert chain of custody. Many systems report the same event at different time intervals. Too fast updates tend to saturate the network. Too slow refresh rates cause the data to be too “stale” for targeting systems. Operational discipline derived from years of military testing and simulation is reused across the entire portfolio of assets / applications.
The Heartbeacon is neutral to middleware – HOWEVER! Not all Middleware is created with the same design / functionality. The Heart Beacon involves gathering state meta data via the heartbeat andPlacing this data (state meta data snapshots) into temporary QUEUES And other temporary structures for onward delivery to the System Administrations (S-6’s) who concatenate this data into heartbeat messages. A change in the platform / user state must be monitored andUpdated and is THRESHOLD sensitive – a user / platform that hasn’t responded is said to be “stale” and shown in the application accordingly
Queues // Thresholds
6
Fares
Crew
Departure
Cargo Reservations
Ticketing
CodeShare
Mechanics
Qmgr - AAA
Qmgr - BBB
Qmgr - CCC
Qmgr - DDD
Only one problem:
This is only 1
server in 1 airport!
Example Application EnvironmentHeart Beacon option: The TCP/IP sub-protocol’s heartbeat function acts as a mini publish-subscribe mechanism that is set to gather state meta data from cloud-subscriber-user at pre-defined intervals by cloud-subscriber-administrator. State meta data is temporarily transferred to files, QUEUES, folders, caches and other temporary data storages awaiting pick up by transport-agent to the cloud-management-broker that aggregates cloud-subscriber-user state management data into heartbeat state management snapshots (REST Representational State Transfer compliant) for onward replication (copy) to other cloud-subscriber-administrators to effect inter and intra cloud infrastructure changes accommodating group adhoc spontaneous integration and cloud-subscriber-GROUP (described by Organization Identification ORG ID) affiliation / de-affiliation based on current mission, operation, scenario goal.
QUEUE MANAGER = Perfect for the Heart Beacon
7
Qmgr - AAA
Qmgr - BBB
Let’s Multiply!
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Qmgr - BBB
Qmgr - DDD
Crew
Mechanics
Crew
Mechanics
Crew
Mechanics
Crew
Mechanics
Qmgr - DDD
Fare Quotes are being processed by all endpoints.
Typically Overlooked:
Are there other apps [Blue
Force Tracker] running on those
endpoints?
Departure
Cargo
Departure
Cargo
Departure
Cargo
Departure
Cargo
Qmgr - AAA
Example Application Environment
Net Centric Warfare Challenge – propagate Sync delta message to update Concept of operations across many networks / subnetsThat are in flux – constantly monitoring end points / platforms / sensors / soldiers
More than 1 type of Blue Force Tracker! Land Warrior/ Grenadier Brat.. Apps for different business units
Avada Confidential - Feedback Monitoring - 2010-11 8
Qmgr - AAA
Qmgr - BBB
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Fares
Ticketing
Qmgr - BBB
Typically Overlooked:
Which endpoints more closely match the PROFILE of travel portal customers or
GEOGRAPHY of portal
customers ?
Departure
Cargo
Departure
Cargo
Departure
Cargo
CodeShare
Reservations
Qmgr - AAA
Example Application Environment
User Profile: Role By OrganizationalType – ORG IDOr type of device:URN: Uniform Resource Name
Each Heartbeat Interval, the geo-location or 10 digitGrid coordinate isSampled. Change in geo-loc isExtrapolated into State – moving, halt, stale, in-op..
9
Fares
Qmgr -EWR
The MQ Admins get ready for a long night because alerts will be firing often based on old traffic patterns. The Biz units get calls wondering why customers are calling in because of slow response times for quotes. Are the servers tuned and optimized? Yes – that is not the issue.
Qmgr - CCC
TravelPortal
Qmgr - DDD
AirlinePortal
Fares
Qmgr - HPN
Fares
Qmgr - LGAEach Unit has its own assumptions of typical business transactions.
Example Application Environment
Beacon Tech / Avada = millisecond response
Disciplined process & temporal sampling discipline protects bandwidth And limited physical resources e.g., server cycles & available time slots Traffic patterns change = network changes
Avada Confidential - Feedback Monitoring - 2010-11
10
Collaborative, Delegated Management
Win2k
OS/390 AIX
AS400
HPUX
Linux
http http
Solaris
HP-nonStop
MQ Adminteam
LGA Qmgrs HPN QmgrsEWR QmgrsJFK Qmgrs ISP-Qmgrs
The MQ Admin has scope to all Qmgrs, Queues, etc. & partial authority to change attributes and settings.
Infrared360 enables role-based visibility into application environments
ROLEBASED
FCC Command &
Control Management
Framework
Heartbeat message#73334 accept / delay
ORG ID / URN
“Rose Colored” Glasses
11
Collaborative, Delegated Management
LGA Qmgrs
Win2kOS/390 AIX
AS400
HPUX
Linux
http http
HPN QmgrsEWR Qmgrs
Solaris
JFK Qmgrs ISP-Qmgrs
HP-nonStop“Invent”
Fares app team = “TF Bling-Bling”
MQ Adminteam
The Fares Biz Unit has scope to only the Fare application Objects, but no authority to modify them.
Infrared 360 enables role-based visibility into application environments
Get / put state meta data from cloud-subscriber-user at pre-defined intervals by cloud-subscriber-administrator. State meta data is temporarily transferred to files, queues, folders, caches awaiting pick up by transport-agent to the cloud-management-broker that aggregates cloud-subscriber-user state management data into heartbeat state management snapshots (REST Representational State Transfer compliant) for onward replication (copy) to other cloud-subscriber-administrators to effect inter and intra cloud infrastructure changes accommodating group adhoc spontaneous integration and cloud-subscriber-GROUP (described by Organization Identification ORG ID) affiliation / de-affiliation based on current mission, operation, scenario goal.
Access to target objects are secured by group and role permissions and audit trails mark every action taken
“The GRAIL”
SC2OPHeartbeat Sync Delta message
# N, N +1 accept / delay
12
Collaborative, Delegated Management
Travel-Portal
LGA Qmgrs JFK Qmgrs
Win2k
OS/390 AIX
AS400
HPUX
Linux
http http
Solaris
HP-nonStop
Fares app team
Travel Portal team
MQ Adminteam
The Travel Portal Biz Unit has scope to only the interface to airline fare quote system.Infrared360 enables role-based visibility into application environments
“Trusted Spaces”
Unified Comms CenterEAC, EOC, DMOC,TOC…
COMMON SYMBOLS
III
THE BIG REDBUTTON
II
SYNCHRONIZEDEVENTS /
ALERTS
I
CAP
TASK FORCE-BLING $
vBET: Network OVERLAY / JBFSA MDL
Integrated System Control:
ISYSCON
UTOAdhocOrgs
Geospatial OVERLAY
GENI NET SNAPSHOT / Heart Beat Message
Easy ConfigReconfig byTask / Mission
NIEMPayload
CAP
OPS / Intel / Log / Civil Affairs.CFO / CEO / Security / Logistics
Visibility7By RoleOnly seeWhat youAllowedTo see
Heartbeat Sync DeltaMessage accept y/n?
13
An Interactive ApproachHow do these biz units collaborate to adjust monitoring thresholds for the new Fares application and associated promo?• Feedback Monitoring: “A practical approach to working with Business Units”
Situational Awareness CDRLS
.0001 .05 .01 .1 5 15 30 99 .0001 .05 .01 .1 5 15 30 99
SYNCSEED_CLOUD
THINK SYNCTHINK SYNC
THINK SYNC
SYNCSEED_CLOUD
SYNCSEED_CLOUD
Single, Authoritative, Reliable blip / event / alert filteredFrom n Situational Awareness systems
just say no to swivel chair
analysis!
Different view from same Common Picture Different view from same Common Picture- Uncleared soldiers on battlefield do not see Intel source
Ops sees OpsIntel sees Intel
A / B /C / D/ n.. -Pentagon- Wall Street- Global NOC
NISTFort Collins, CO
-IEPD -- Naming Design Rules NDR- Conformance Tool- Payload / data island to Common Alert Protocol
IRODS: Rules Repository / Engine across n domains
CDX NIEM TECHIn The Cloud
Infrared360™
CI-BER: CyberInfrastructure for Billions of Electronic Records
NARA / NCASTUNC CHAPEL HILL
Avada Confidential - Feedback Monitoring - 2010-11 14
Monitoring Scenarios are usually based upon:
Immediate Thresholds – historical assumptions of current SLAs (MQ Admins generally set these)
Statistical Measurements – capacity teams for each biz unit : MQ Admins, fares app. team, portal app. team)
From historical trends ¹
From test harnesses ²
From transaction measuring software ³
Avada Confidential - Feedback Monitoring - 2010-11 15
Current Methods
Patterns are usually based on historical volumes:
Assuming the same rate of business ... when your biz wants better?
Cost of analysis typically low = cost of error typically high!
Patterns are usually based upon previous offerings
But expectations are now based upon market condition? – different volume of users, the time of the offering?
If so, the previous patterns may not apply.
Avada Confidential - Feedback Monitoring - 2010-11 16
Feedback Monitoring:
Deriving Patterns and Behavior
Sometimes there are no previous base statistics so behavior patterns are not yet determined
Even more need for collaborative intermediate analysis• Pattern detection• Application behavior• Peaks & Valleys• Volumes• Timeframes
Avada Confidential - Feedback Monitoring - 2010-11 17
Supporting Agile Business
Corporations are in difficult economic conditions and/or competitive markets
Need for creativity and agility to create new offerings in order to compete
New offerings are business decisions
Outside of IT
Regardless of IT
Are time dependent to get market share
Are based completely on promised return on investment
Avada Confidential - Feedback Monitoring - 2010-11 18
Market Driven Examples:
Travel – special fare offerings
Banking – special rates accounts
Insurance – competitive rate offerings
Retail – special price + limited time offerings
Utilities – special rate offerings
Manufacturing – limited resource availability
All of the above – new partners, inventory, financial incentives &
liabilities
Avada Confidential - Feedback Monitoring - 2010-11 19
Problem Area for Threshold Planning
Issue: Historical basis not yet formed Threshold basis not yet determined Market conditions not certain Success of acceptance of new business app
Goal: Early pattern and behavioural recognition
Avada Confidential - Feedback Monitoring - 2010-11 20
Best Practices for Setting Thresholds
Modelling New Business Applications • Use real data! “hello world” is not a transaction
• Model transactions in test/QA/etc using the same payloads
as would be in Production
• Model them while running other transactions volumes on
the same Qmgr, other queues, as would be in Production
• Monitor the transactions in test/QA/etc the same as you
would in Production (better to know it now than later)
Avada Confidential - Feedback Monitoring - 2010-11 21
Modelling new business applications
Review findings with business units
Use comparatives – is this normal, not normal?• Fare app Trend Chart example
Offer opinions on ability to run & support • What are alternatives if predictions are wrong?
Analyze real peaks in application trends• seasonal, weekly, daily (peak levels for each)• your thresholds are built for the peaks – not the average!
Create 'stepped' warning levels for MQ Admins • at threshold A, at threshold B, etc?
Best Practices for Setting Thresholds
Avada Confidential - Feedback Monitoring - 2010-11 22
Modelling New Business Applications • Review expectations and tactics with business units.• Gather statistics over short intervals during initial rollout.• Determine and set threshold values to watch – initial hours.• Gather statistics over medium intervals during initial rollout.• Determine and set threshold values for comfortable limits for
that day (8am-11am, 11am-2pm, 2pm-5pm, 5pm-8pm).• Gather statistics over longer intervals during initial rollout.• Determine and set thresholds values based upon daily totals, then based upon each daily pattern for that week.• Monitor thresholds for THIS APP should now be in place.
• Use same feedback loop in early production monitoring for each unique application !!!
Best Practices for Setting Thresholds
Avada Confidential - Feedback Monitoring - 2010-11 23
Infrared360 Provides:
• Collaborative visibility to business applications environments• Short, mid, long term threshold alerting• Real-time middleware object detail statistics visibility• Comparative statistical charting of real-time middleware objects• Real-time administrative interface to problem area for quick problem resolution • Self-healing capabilities ties alerts to pushed Admin changes• Automated replay and volume testing of real transactions• Instant visibility to application environment without pre-set monitoring infrastructure
for Enterprise Middleware Environments
Infrared360 Value Proposition
Avada Confidential - Feedback Monitoring - 2010-11 24
Feedback Monitoring Best Practices
Thank you!