system architecture design exercises may 2012

System Architecture Design

Exercises May 2012

Copyright©2012 Esri All

rights reserved.

Course version 10.0. Version release date January 2012.

Printed in the United States of America.

The information contained in this document is the exclusive property of Esri. This work is protected under United States copyright law and other international copyright treaties and conventions. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage or retrieval system, except as expressly permitted in writing by Esri. All requests should be sent to Attention: Contracts and Legal Services Manager, Esri, 380 New York Street, Redlands, CA 92373-8100 USA.

EXPORT NOTICE: Use of these Materials is subject to U.S. export control laws and regulations including the U.S. Department of Commerce Export Administration Regulations (EAR). Diversion of these Materials contrary to U.S. law is prohibited.

The information contained in this document is subject to change without notice.

U. S. GOVERNMENT RESTRICTED/LIMITED RIGHTS

Any software, documentation, and/or data delivered hereunder is subject to the terms of the License Agreement. The commercial license rights in the License Agreement strictly govern Licensee's use, reproduction, or disclosure of the software, data, and documentation. In no event shall the U.S. Government acquire greater than RESTRICTED/LIMITED RIGHTS. At a minimum, use, duplication, or disclosure by the U.S. Government is subject to restrictions as set forth in FAR §52.227-14 Alternates I, II, and III (DEC 2007); FAR §52.227-19(b) (DEC 2007) and/or FAR §12.211/12.212 (Commercial Technical Data/Computer Software); and DFARS §252.227-7015 (NOV 1995) (Technical Data) and/or DFARS §227.7202 (Computer Software), as applicable. Contractor/Manufacturer is Esri, 380 New York Street, Redlands, CA 92373-8100, USA.

@esri.com, 3D Analyst, ACORN, Address Coder, ADF, AML, ArcAtlas, ArcCAD, ArcCatalog, ArcCOGO, ArcData, ArcDoc, ArcEdit, ArcEditor, ArcEurope, ArcExplorer, ArcExpress, ArcGIS, ArcGlobe, ArcGrid, ArcIMS, ARC/INFO, ArcInfo, ArcInfo Librarian, ArcLessons, ArcLocation, ArcLogistics, ArcMap, ArcNetwork, ArcNews, ArcObjects, ArcOpen, ArcPad, ArcPlot, ArcPress, ArcPy, ArcReader, ArcScan, ArcScene, ArcSchool, ArcScripts, ArcSDE, ArcSdl, ArcSketch, ArcStorm, ArcSurvey, ArcTIN, ArcToolbox, ArcTools, ArcUSA, ArcUser, ArcView, ArcVoyager, ArcWatch, ArcWeb, ArcWorld, ArcXML, Atlas GIS, AtlasWare, Avenue, BAO, Business Analyst, Business Analyst Online, BusinessMAP, CommunityInfo, Database Integrator, DBI Kit, EDN, Esri, Esri—Team GIS, Esri—The GIS Company, Esri—The GIS People, Esri—The GIS Software Leader, FormEdit, GeoCollector, Geographic Design System, Geography Matters, Geography Network, GIS by Esri, GIS Day, GIS for Everyone, GISData Server, JTX, MapIt, Maplex, MapObjects, MapStudio, ModelBuilder, MOLE, MPS—Atlas, PLTS, Rent-a-Tech, SDE, SML, Sourcebook·America, Spatial Database Engine, StreetMap, Tapestry, the ARC/INFO logo, the ArcGIS logo, the ArcGIS Explorer logo, the ArcPad logo,the Esri globe logo, the Esri Press logo, the GIS Day logo, the MapIt logo, The Geographic Advantage, The Geographic Approach, The World's Leading Desktop GIS, Water Writes, www.esri.com, www.geographynetwork.com, www.gis.com, www.gisday.com, and Your Personal Geographic Information System are trademarks, registered trademarks, or service marks of Esri in the United States, the European Community, or certain other jurisdictions.

Other companies and products mentioned herein may be trademarks or registered trademarks of their respective trademark owners.

mailto:@esri.com�

http://www.geographynetwork.com/�

http://www.geographynetwork.com/�

Table of Contents Class Exercise (Questions) Page

1.0 System Design Process ........................................................... 1-1 2.0 GIS Software Technology ....................................................... 2-1 3.0 GIS Software Performance .................................................... 3-1 4.0 GIS Server Performance ........................................................ 4-1 5.0 GIS Data Architecture ............................................................ 5-1 6.0 Network Communications...................................................... 6-1 7.0 GIS Product Architecture ...................................................... 7-1 8.0 Platform Performance ............................................................ 8-1 9.0 Information Security .............................................................. 9-1 10.0 Performance Fundamentals ................................................. 10-1 11.0 City of Rome .......................................................................... 11-1

Class Exercise Answer Sheet

ESRI | 0BSystem Design Process 1-1

Lesson 1: System Design Process

Purpose: A common question asked when purchasing a server is “what can this hardware do?” The purpose of this exercise is to use the CPT Platform Capacity Calculator tool to evaluate the best Esri hardware bundle selection for your server environment. The exercise provides a list of hardware platform bundles for you to evaluate. Price and GIS software licensing for each bundle is the same, while the host platform for each vendor is different. You will use the Platform Capacity Calculator tool to compare the performance output for each bundled solution.

Scenario: Your company has approved funding to purchase new hardware for your upcoming ArcGIS Server deployment. You have decided to review the Esri Hardware Partner Promotional Offers available on the Esri Web site (http://www.esri.com/partners/apps/hw_promo/index.cfm). You are interested in purchasing a Server with an entry level Enterprise Advanced ArcGIS Server license. You will use the Platform Capacity Calculator to identify performance and capacity for the available server candidates. A total of 14 ArcGIS Server hardware bundle solutions are available on the Esri Web site. You are interested in bundles that include an ArcGIS Server Advanced Enterprise license, which narrows your selection down to three vendor candidates. The cost for these three promotion bundles are the same. You are looking for a 4-core server (entry level software license) with at least 12 GB of memory (3 GB per server core). All of the server candidates satisfy your licensed core and memory requirements. You can use the Platform Capacity Calculator to evaluate performance capacity and speed for each of these servers. Performance of the server is determined by the processer. Following list identifies the processor candidates for each of these servers.

• Dell PowerEdge R610, Intel Xeon X5687 4 core (1 chip) 3600 MHz processor. • HP ProLiant DL360, Intel Xeon X5672 4 core (1 chip) 3200 MHz processor. • IBM System x3550, Intel Xeon E5640 4 core (1 chip) 2666 MHz processor.

Exercise Goals: In this exercise, you will use the CPT Platform Capacity Calculator to review performance of three hardware bundles.

• You will use performance metrics from your current operational environment to establish a performance baseline for your existing ArcGIS for Server REST mapping services.

• You will then evaluate peak capacity (throughput and peak concurrent users) for each of the offered bundle configurations.

• You will then evaluate impact of user productivity on your peak concurrent user estimates. • Finally you will evaluate deployment options comparing peak throughput for a 4-core physical host server

configuration with two optional virtual server configurations (2x 2-core and 1x 4-core). Best practice: When using the Capacity Planning Tool (CPT), it is good to make a new copy for use in your system design analysis (keep the original as a master). Save a copy of the CPT and extend the file name by including a prefix representing your project effort (i.e. ExerciseArc11CapacityPlanning0100.xlsx). It is good to include the CPT version number (Arc11CapacityPlanning0100) for future reference.


Step 1: What is my current hardware (X5160 4 core) throughput capacity (TPH)? Current AGS10 REST MXD SDE peak throughput is 17,000 TPH. Procedure:

• Use Platform Capacity Calculator located on the Hardware tab to complete your analysis. Best practice: Copy of each Platform Capacity analysis can be saved for future reference by using the Excel Move or Copy sheet command. Use copy sheet command (check create a copy block) to create new Hardware tab copy and select move to end. Label new tab 1.1-2 (Lesson 1 Questions 1-2).

• Select your current platform configuration in column A. • Select TPH as the capacity output in column C. • Select Physical platform in column E.

1. What is the capacity of the Xeon 5160 4 core (2 chip) 3000 MHz platform? a. 11,400-22,800 for an AGS10 REST MXD SDE service. b. 18,600 – 37,200 for an AGS10 REST MSD SDE service. c. 11,800 – 23,600 for an AGS10 WMS MXD SDE service. d. All of the above are true.

Note: The MSD optimized rendering engine doubles the server capacity. Recommended use of the optimized MSD rendering engine will be discussed in Lesson 3. The ArcGIS 10.1 release will use the MSD rendering engine for all Web mapping services (MXD will not be supported).

2. What is the complexity of your current Web applications? a. Light b. Medium Light c. Medium d. Medium Heavy


Note: Map service complexity depends on several factors, including number of display layers and dynamic functions executed with each display. The Platform Capacity Calculator shows a performance range from light to medium complexity for each workflow. Display complexity will be discussed in Lesson 3.

Step 2: What is the throughput capacity (TPH) of the new hardware bundles? Procedure:

• Use Platform Capacity Calculator located on the Hardware tab to complete your analysis. Best practice: Use Excel move or copy sheet command to create a copy and move to end for each platform analysis. Label new tab 1.3 (Lesson 1 Question 3).

• Select TPH as the capacity output in column C. • Select Physical platform in column E. • Select platform configuration for each of the candidate platforms in column A.

Note: You can save multiple copies of Excel tab 1.3 and label 1.4 and 1.5.

3. What is the ArcGIS 10 REST MSD capacity of the Dell Xeon X5687 4 core hardware bundle? a. 38,800 – 77-600 TPH b. 50,000 – 100,000 TPH c. 53,300 – 106,600 TPH d. 69,500 – 139,000 TPH

Note: Reset platform selection (column A) for the following answers.

4. What is the ArcGIS 10 REST MSD capacity of the HP Xeon X5672 4 core hardware bundle? a. 38,800 – 77-600 TPH b. 50,000 – 100,000 TPH c. 53,300 – 106,600 TPH d. 69,500 – 139,000 TPH


5. What is the ArcGIS 10 REST MSD capacity of the IBM Xeon E5640 4 core hardware bundle? a. 38,800 – 77-600 TPH b. 50,000 – 100,000 TPH c. 53,300 – 106,600 TPH d. 69,500 – 139,000 TPH

Step 3: How many concurrent users can the new hardware bundles support? Procedure:

• Use Platform Capacity Calculator located on the Hardware tab to complete your analysis. Best practice: Use Excel move or copy sheet command to create a copy and move to end for each platform analysis. Label new tab 1.6.

• Select Users as the capacity output in column C. • Select Physical platform in column E. • Select platform configurations for each of the candidate platforms in column A and label new tabs 1.6, 1.8,

and 1.9.

6. What is the ArcGIS 10 REST MSD capacity of the Dell Xeon X5687 4 core hardware bundle? a. 193 – 386 users b. 148 – 296 users c. 139 – 278 Users d. 108 – 215 Users

Note: Reset platform selection (column A) for the following answers.

7. What is the ArcGIS 10 REST MSD capacity of the HP Xeon X5672 4 core hardware bundle? a. 193 – 386 users b. 148 – 296 users c. 139 – 278 Users d. 108 – 215 Users


8. What is the ArcGIS 10 REST MSD capacity of the IBM Xeon E5640 4 core hardware bundle? a. 193 – 386 users b. 148 – 296 users c. 139 – 278 Users d. 108 – 215 Users

9. Which server processor supports the highest number of users? a. Dell PowerEdge R610, Intel Xeon X5687 4 core (1 chip) 3600 MHz processor. b. HP ProLiant DL360, Intel Xeon X5672 4 core (1 chip) 3200 MHz processor. c. IBM System x3550, Intel Xeon E5640 4 core (1 chip) 2666 MHz processor. d. All hardware bundles support the same number of users.

Step 4: How does my use productivity impact concurrent user capacity? Procedure:

• Use Platform Capacity Calculator located on the Hardware tab to complete your analysis. Best practice: Use Excel move or copy sheet command to create a copy and move to end for each platform analysis. Label new tabs 1.10, 1.11, 1.12, and 1.13.

• Select Users as the capacity output in column C. • Select Physical platform in column E. • Select Dell platform configuration in column A. • Enter special productivity in column G to complete analysis.

Note: CPT assumes default productivity when no entry in column G.

10. What are the AGS10 REST MSD peak users with productivity at 6 displays per minute? a. 444 – 888 users b. 222 – 444 users c. 148 – 296 users d. 89 – 178 users




13. How does user productivity impact peak concurrent users? a. Lower user productivity reduces peak number of supported users. b. Lower user productivity increases the minimum number of supported users c. Higher user productivity reduces peak number of supported users. d. Higher user productivity increases peak number of supported users.

Warning: Peak number of supported users will depend on user productivity and map display complexity.


Step 5: What is my optimum deployment strategy if I choose to use virtual servers? Procedure:

• Use Platform Capacity Calculator located on the Hardware tab to complete your analysis. Best practice: Use Excel move or copy sheet command to create a copy and move to end for each platform analysis. Label new tab 1.14, 1.15, 1.16, and 1.17.

• Identify best performance configuration when deploying ArcGIS for Server in a virtual server environment:

o Select Dell X5687 4-core host platform configuration in column A. o Select TPH as the capacity output in column C. o Select VMware Server platform in column E. o Enter virtual core/node in column F:G. o Enter number of virtual servers in column H.

14. What is the AGS10 REST MSD capacity (TPH) for a single 4-core physical server machine? a. 44,400 – 88,800 TPH b. 38,100 – 76,200 TPH c. 53,300 – 106,600 TPH d. 48,500 – 97,000 TPH

Note: Select Physical platform selection in column E to generate physical server results.


15. What is the AGS10 REST MSD capacity (TPH) for two 2-core Virtual Server machines? a. 44,400 – 88,800 TPH b. 38,100 – 76,200 TPH c. 50,000 – 106,600 TPH d. 48,500 – 97,000 TPH

Note: Select VMware platform selection in column E, identify the number of virtual server core/node in column F:G, and identify number of virtual servers in column H (total number of core = 4).

16. What is the AGS10 REST MSD capacity (TPH) for one 4-core Virtual Server machine? a. 44,400 – 88,800 TPH b. 38,100 – 76,200 TPH c. 50,000 – 106,600 TPH d. 48,500 – 97,000 TPH

17. Which Virtual Server configuration provides the higher peak throughput? a. Two 2-core virtual server machines. b. One 4-core virtual server machine. c. Both configurations have the same throughput. d. None of the above.


Lesson 1 Review Questions

1. What is the primary objective of the System Architecture Design process? a. Best practices for configuring GIS systems b. Understanding platform performance and scalability. c. Methods for test and evaluation of GIS system performance. d. Establishing platform and network requirements based on user workflow needs

2. Why is System Architecture Design important? a. System capacity is limited by the weakest component. b. Identifying required component specifications early can save money. c. Building a balanced system design improves performance. d. All of the above.

3. How can we manage implementation risk? a. Test the production system to identify performance bottlenecks. b. Purchase the best hardware technology. c. Establish performance targets and validate implementation milestone compliance. d. Maintain good vender relationships.

4. What is the recommended best practice when completing a system architecture design planning process?

a. Use the capacity planning tool to support an integrated business needs assessment. b. Complete a System Architecture Design strategic plan after the user needs assessment. c. Ask the hardware vendor to provide a system sizing assessment. d. Follow sizing recommendations provided in the software documentation.

5. When is it important to validate workflow performance compliance? a. Evaluate performance compliance during initial prototype testing. b. Validate system performance goals are met during initial production deployment. c. Establish periodic performance validation milestones throughout production. d. All of the above.

6. How can capacity planning tools reduce deployment risk? a. Identify hardware required to satisfy peak user business needs. b. Establish reasonable performance expectations for software technology selection. c. Establish reasonable workflow performance targets for performance test and validation. d. All of the above.

ESRI | 1BGIS Software Technology 2-1

Lesson 2: GIS Software Technology

Purpose: The first step in every design process is to identify your user requirements. User types are then grouped together and assigned to different software technology patterns or use cases based on their work activity. Common technology patterns are represented in the CPT as Standard Workflows, providing a baseline for selecting appropriate workflow performance targets for your design. These Standard Workflows include software processing times (service times) for components installed on the system. Learning how to select appropriate workflow performance targets to represent your GIS operations is the first step in completing your system design.

Scenario: The CPT Calculator is used to generate Standard Workflows identified on the CPT Workflow tab. The workflow nomenclature (recipe) identifies the selected software technology performance factors. The performance factors are used by the CPT Calculator to establish workflow performance targets for the selected software technology.

• ArcGIS for Server (AGS) or ArcGIS for Desktop (AGD) software technology • Map Document format (MXD or MSD) or Imagery (MosaicDS, RasterDS) • Display density (V is for vector only, R includes imagery layer)/ Imagery (processed or dynamic) • Percent of display from map cache (recipe shows percent of display dynamic). • Display Complexity (light, medium light, medium, medium heavy, heavy, 2x Medium, 3x Medium) • Display resolution (map size rounded off to hundreds of pixels) • Output format (JPG, PNG8, PNG24, PNG32, PDF, Feature, ICA)

Exercise Goals: In this exercise, you will use the Standard Workflows provided on the CPT Workflow tab to identify your project workflow performance targets.

• The first part of the exercise will match the CPT Recipe for each project workflow with the appropriate software technology performance factors (software design specifications).

• Next you will select each of the project workflows from the list of Standard Workflows in the CPT Workflow tab, and include a copy of each workflow in your Project Workflows section.

• Finally, you will use your selected project workflows on the CPT Design tab to complete your system architecture design.


Step 1: What are my business requirements? The first step in the system architecture design process is to select the right GIS software technology workflow performance specifications required to satisfy your user business requirements. These software technology patterns are represented in the CPT as user workflows. During your business needs assessment you identify four primary GIS use cases that you wish to include in your system design. These include GIS desktop editors and viewers located on your local network environment, some remote GIS desktop viewers located at remote sites connected over your local area network, and Web mapping services used by internal LAN and WAN clients and by public clients over the Data Center Internet gateway. You have discussed these business needs with your user community and selected the following Standard Workflows to represent each of the identified use cases in your design. The desktop editors will be connecting to the maintenance geodatabase (DBMS) and performing standard edit functions. A medium complexity workflow was selected to represent system loads for standard functions used in the editing workflow. Nickname: DeskEdit

1. What is the software technology used for the DeskEdit workflow? a. ArcGIS 10 for Desktop. b. ArcGIS 10 for Server. c. ArcGIS 9.3 for Desktop. d. ArcGIS 9.3 for Server.

The desktop viewers will be using simple focused map documents optimized for quick display. A more optimistic light complexity desktop workflow was selected to represent the reduced system loads. Nickname: DeskView

2. What is the display resolution used for the DeskView workflow? a. 400x600 pixels. b. 600x800 pixels. c. 1024x768 pixels. d. 1280x1024 pixels.


The Web mapping workflow will represent a variety of map services for both internal and public users. A medium Standard Workflow was used to provide a conservative performance target for these services. Nickname: WebMap

3. What is the map document used for the WebMap workflow? a. MXD b. MSD c. Map display d. Map output

The remote GIS desktop viewers will be using simple focused map documents optimized for quick display. A Citrix workflow will be used for all remote clients for optimum performance over limited network bandwidth connections. A more optimistic light complexity desktop workflow was selected to represent the reduced system loads. Nickname: CitrixView

4. What is the map density used for the CitrixView workflow? a. Raster image b. Vector only c. Real image d. Virtual map

Step 2: What are my workflow performance targets? Use the Standard Workflows from the CPT Workflow tab to configure the Project Workflows identified in Step 1 above.

Note: Standard Workflows and Project Workflow section are both located on the CPT Workflow tab.

Procedure: • Locate each Standard Workflow (match the workflow recipe). • Select the Standard Workflow you wish to include in your Project Workflow section (Select and

copy workflow row). Best Practice: Use your mouse right button to access CPT Excel functions. The mouse selection menu is most common across the supported CPT Excel software releases.

• Insert copied row into Project Workflow section (Insert copies cells) • Add Nickname as suffix for each workflow recipe (follow nickname with underscore).


CPT Workflow tab Project Workflows.


Once you complete your Project Workflow section, use the total software service time (column N) to check you work.

5. What is DeskEdit total software service time? a. 0.238 b. 0.248 c. 0.278 d. 0.475

6. What is DeskView total software service time? a. 0.238 b. 0.248 c. 0.278 d. 0.475

7. What is WebApp total software service time? a. 0.238 b. 0.248 c. 0.278 d. 0.475

8. What is CitrixView total software service time? a. 0.238 b. 0.248 c. 0.278 d. 0.475

Step 3: How do I use my selected project workflows to complete my system design? During your business needs assessment you identified a total of 4 workflows you need to complete your design. You selected these workflows based on the software technology pattern and the complexity of GIS information products required by the user community. Next you need to identify where these users are located, identify the peak throughput loads for each workflow, and select the network bandwidth and platform architecture used to complete your system design. The CPT Design tab can be used to complete the system design analysis and identify the final platform configuration.

During your business needs assessment you compiled the following information on user location and peak usage for each project workflow. User network location and available bandwidth is provided in Columns A and B. Peak usage for each user workflow is identified in columns C through E. You will use this information to complete your requirements analysis on the CPT Design tab.


Business Requirements Summary

Note: ArcGIS for Desktop DeskEdit workflows are for users located on the LAN and CitrixView

workflows are for users located on the WAN.

Once your project workflows are identified in the workflow tab, the CPT Design requirements module can be configured to complete the system architecture design. Workflows are identified based on user network location (3 LAN workflows, 2 WAN workflows, and 1 Internet workflow). Available bandwidth is identified for each network connection. Platform hardware is selected on the platform tier. Network suitability and system design analysis are completed by the Excel application. Hardware solution is provided once the CPT Design tab is properly configured. CPT Design configuration

Note: The CPT Design1 tab is preconfigured for your use during this exercise. Use the Design1 tab for step a and skip to step b when using Design1 to complete this exercise.


a) Use the Excel Move or Copy command to create a copy of the CPT Design tab. Locate the new Design tab to the right of the workbook tabs. You can rename the new design tab for future reference.

Best practice: Copies of each design state can be saved for future reference by using the Excel Move or Copy sheet command. Note: You can skip to Step b if you used the preconfigured Design1 tab.

For this design we will focus on the central data center configuration. • Select and delete the CPT Design tab GREEN rows within the Requirements Analysis module that

represent remote site locations. Warning: Do not delete the GREEN row at the bottom of the Requirement Analysis module. This GREEN

row is a template we will use later for adding new remote sites. The data center has LAN, WAN, and Internet network connections. Peak user workflows are configured based on user location. Available bandwidth is identified for each network connection. Platform hardware is selected on the platform tier. Network suitability and system design analysis are completed by the Capacity Planning tool. Hardware solution is provided once the CPT Design tab is properly configured.

Note: We will discuss remote site configuration strategies in Lesson 6.

b) Configure the Requirements Analysis section of the new CPT Design tab to identify user locations. Best practice: To delete rows, select complete row (mouse located on row number during selection) and then delete the row. To add rows, select and copy one of the existing workflow rows and then select row and insert copied cells where you wish to add the new workflow.

• Three user workflows are located on the LAN network connection (between the LAN and WAN rows). • Two user workflows are located at remote sites on the WAN network connection (between the WAN and

Internet rows). • One workflow provides services for clients over the Internet network connection (below the Internet row). • Select the appropriate workflow from your Project Workflow list using the dropdown menus located in the

column B white cells. • Enter the peak throughput (users or TPH) for each of the workflows as identified in the business

requirements summary. Input users in column C and TPH in column D. • Delete any of the rows you do not need to complete your design.

Warning: Be sure and select and delete the complete row.

c) Enter the peak usage requirements for each workflow in the white cells in column C and D.

9. What is the total throughput identified at the bottom of the user column? a. 50 users b. 75 users c. 120 users d. 195 users


10. What is the total throughput identified at the bottom of the services column? a. 25,000 TPH b. 50,000 TPH c. 75,000 TPH d. 100,000 TPH

Warning: Be careful not to change any of the formulas located in CPT colored cells. These formulas are used by Excel to complete the design analysis.

d) Enter data center bandwidth connections in the column H white cells for each network row. • LAN = 1000 Mbps • WAN = 45 Mbps • Internet = 45 Mbps

11. What is the peak utilization identified for the Internet row in column I? a. 4 percent b. 46 percent c. 62 percent d. 78 percent

e) Complete the CPT Design hardware platform selection in column B white cells for the Client and hardware platform tier. • Intel Core i3-2120 2 core (1 chip) 3100 MHz workstation for the Client. • Xeon X5687 4 core (1 chip) 3600 MHz servers for each platform tier.

Note: We will discuss platform architecture in Lesson 7 and platform selection in Lesson 8.

f) Identify the recommended platform solution.

12. WTS Citrix platform tier requires ____ servers with peak utilization of ____ percent. a. 2 servers, 47.6 percent b. 2 servers, 46.3 percent c. 1 server, 31.0 percent d. Solution does not include Citrix servers.

13. GIS platform tier requires ____ servers with peak utilization of ____ percent. a. 2 servers, 47.6 percent b. 2 servers, 46.3 percent c. 1 server, 31.0 percent d. Solution does not include GIS servers.

14. DBMS platform tier requires ____ server with peak utilization of ____ percent. a. 2 servers, 47.6 percent b. 2 servers, 46.3 percent c. 1 server, 31.0 percent d. Solution does not include DBMS servers.


Best practice: The SDE Geodatabase platform is a scale up environment and normally must be configured on a single DBMS server. Only active production server nodes are shown on the CPT Design tab - failover servers are not included in the display.

Warning: Selection of medium complexity Standard Workflow does not guarantee design compliance. Heavy GIS analysis functions and complex data models (high number of map layers, features, and data model dependencies can generate much heavier workflow loads). Workflow design best practices discussed in Lesson 3 are important considerations in building a GIS that operates within medium complexity performance standards. Light complexity workflows require careful attention to design specifications and project management (performance validation) throughout system implementation to deliver within these much tighter performance goals.


Lesson 2 review questions

1. What is the optimum software technology for GIS power users? a. ArcGIS for Desktop. b. ArcGIS Server Web Applications c. ArcGIS Engine Application d. ArcGIS Mobile ADF Application

2. What is the optimum software technology for casual Web users? a. ArcGIS Desktop b. ArcGIS Server c. ArcGIS Engine d. ArcGIS Mobile

3. What is fundamental about GIS that encourages people to share their work? a. Benefits provided by common understanding of map display results. b. Benefits provided by shared hardware platform environments. c. Benefits gained by sharing geospatial data resources. d. Benefits gained by sharing standard data models and business rules.

4. What community benefits most from a centralized Architecture? a. GIS Power users. b. GIS General Users. c. GIS Business Partners. d. IT Administrators.

5. What was the user performance experience moving from ARC/INFO 7.3 to ArcGIS Desktop 8.0 a. User display performance improved by 10% due to more efficient program code. b. User interface contributed to reduced productivity. c. Slower ArcGIS Desktop display performance requires twice the processing time. d. Faster ArcGIS Desktop display performance required half the processing time.

ESRI | 2BSoftware Performance 3-1

Lesson 3: Software Performance

Purpose: There are many ways customers use GIS technology, and every user workflow can be slightly different. The software technology performance factors used in the CPT are selected based on Esri experience in testing and tuning GIS operations. These key performance factors are used by the CPT Calculator to generate custom workflow performance targets that you can use in your project design.

Scenario: During your user needs analysis, you must select a set of workflows for use in your project design. This lesson introduced the CPT Calculator as a tool that allows you to identify some key software specifications (software technology performance factors) and have the Calculator generate appropriate performance targets. You may also want to review the impact of these workflow loads on your selected platform and network environment before making final software technology decisions. System performance and scalability, hardware resources, and software licensing will be key parameters contributing to your technology selection.

Exercise Goals: In this exercise, you will

• Use the CPT Calculator to generate a custom workflow. • Use the CPT Calculator to evaluate system design benefits of the following performance factors:

o Optimized map rendering engine (MSD) o Cached map services o Map display resolution

• Use the CPT Calculator to review hardware performance requirements and software licensing. • Add a new custom workflow in your list of project workflows located on your CPT Workflow tab.


Step 1: How can I use the CPT Calculator to help select the right software performance targets?

a) Select the following Software technology performance factors. Procedure: • Use the CPT Calculator tab to generate performance targets for the following workflow recipe.

Workflow recipe

1. What is the map service output format? a. PNG b. JPEG c. PDF d. Feature

b) Enter your Web mapping user requirements and data source. Web mapping business requirements include 70 local users at 6 displays per minute/client (25,200 TPH) and an additional 50,000 TPH for public clients (total Web mapping throughput of 75,200 TPH).

Note: these were the Web mapping business requirements we used in Lesson 2.

• Enter peak throughput and select SDE_DBMS data source. User workflow requirements and data source

c) Select the following hardware architecture and network settings. Selected hardware architecture and network bandwidth

• Select platform architecture (2 tier, minimum, DC) in column A. • Select hardware platforms (column D:F) and rollover setting (column G).


• Identify remote site (rows 13-14) peak users (column D), network latency (column E), and bandwidth (column G). Local client bandwidth should be 100 Mbps (Workstation NIC bandwidth)

• Select client desktop platform (column C:F).

d) Deploy using the Dell hardware bundle solution identified in Lesson 1. Once you enter your software technology performance factors, user requirements, and hardware architecture the CPT Calculator completes the system architecture design. CPT Calculator Configuration

2. What is the remote site 1 display response time? a. 0.59 sec b. 1.35 sec c. 1.56 sec d. 1.34 sec


3. Identify the number of required ArcGIS Server licensed core (Web/SOC tier). a. 2-core b. 4-core c. 8-core d. 12-core

Best practice: Use Excel move or copy sheet command to create a copy and move to end for each platform analysis. Label new tab 3.4-5.

Step 2: Will optimized map document (MSD) improve performance? • Use the same workflow from Step 1 and change the MapDoc selection (G3) to MSD.




Step 3: What are the advantages in caching base layers and controlling display resolution? • Change the software technology performance factors to the following recipe.

Adjusted workflow description (recipe)

Note: The +mapcache selection in K5 will deliver cached map tiles to the Web client display. This requires negligible server processing, with additional average display traffic of 0.5 Mbpd.

CPT Calculator configuration

6. What is the remote site 1 remote user display response time? a. 0.59 sec b. 1.35 sec c. 1.56 sec d. 1.34 sec



Step 4: How can I add this custom WebLite workflow to my project workflows? Procedure: • Select and copy one of your existing Project Workflow rows and "insert copied cells" to add a new

workflow. Note: New workflow row will be used as a template for adding custom workflow values as a new Project Workflow.

• Locate the Calculator workflow section at the bottom of the CPT Workflow tab.

Note: New custom CPT workflows must first be configured on the Calculator tab before they show up Calculator Workflows on the workflow tab.

• Select and copy blue Calculator Workflow row cells (Columns A through M). • Select the new project workflow template cell in Column A and Paste Special/Values to insert the new

Calculator generated workflow service times into the new project workflow. Best practice: It is good practice to copy the workflow recipe (column A) and Paste Special/Values into the workflow description column (located in column AJ) to document workflow source.

• Insert Nickname (WebLite_) as the new workflow prefix. •

CPT Workflow Tab Project Workflows


8. What is the total Arc11 baseline processing time (column N) for the WebLite workflow? a. 0.59 sec b. 1.35 sec c. 1.56 sec d. 1.34 sec

Warning: Selection of custom workflow performance targets does not guarantee design compliance. Workflow design best practices are important considerations in building a GIS that operates within established performance standards. Attention to design specifications and project management (performance validation) throughout system implementation is necessary to deliver within established performance goals.



1. How can a GIS user author a high performance map display? a. Only show relevant data. b. Use scale dependencies. c. Use integer fields for symbol values. d. All of the above.

2. What are map display functions that hurt performance? a. Use Esri Optimized style for lines and polygons. b. Use annotation instead of labels. c. Use cartographic lines and polygon outline. d. Use indexed fields for text and labeling.

3. How does ArcGIS Desktop build a map display? a. Request spatial features sequentially rendering one layer at a time for display. b. Collect all spatial features from data source then organize on desktop for display. c. Request data source to organize data and transport view to desktop for display. d. ArcObjects are used to render display from database attributes.

4. How may twice the number of layers and features impact map display processing time? a. Increasing number of display layers has minimum impact on display processing time. b. Double user response time with minimal increase in display service time. c. Double display service time and reduce platform capacity by 50 percent. d. Double network traffic with minimum impact on service time and system capacity.

5. How does display output format impact client network traffic? a. PNG24 provides the best compression with raster imagery. b. PDF provides the smaller display traffic for all data types. c. JPEG provides relatively consistent light display traffic for raster and vector imagery. d. Image data source selection has minimum impact on network traffic.

6. What is a primary advantage of using map cache data source? a. Improved geospatial data integration. b. Reduced client network display traffic requirements c. Reduced data administration requirements. d. Reduced display processing requirements.

7. What is the performance impact when using a high quality pre-processed cached cartographic data source?

a. Slower user response time due to quality of the display. b. Higher network traffic due to quality of the display. c. Display performance is roughly the same as a simple cached data source. d. Pre-cache processing timelines will be shorter due to quality of the data.

ESRI | 3BGIS Server Performance 4-1

Lesson 4: GIS Server Performance

Purpose: Proper Web service instance settings are important in leveraging available platform computing resources. This exercise will use the CPT Design tab to demonstrate the importance of proper Web service configurations.

Scenario: Web services should be configured based on expected use patterns. Batch and geoprocessing tasks consume a core processor during execution, while lighter Web services can require 3-5 service instances per core processor to reach peak platform throughput levels. The CPT Design tab can be configured to demonstrate the processing loads for each of these service configurations.


• Configure the CPT Design tab for testing a single BATCH process. • Demonstrate maximum throughput for 1, 2, 3, 4, 5, and 6 service instances on a 4-core server. • Configure the CPT Design tab for testing a single dynamic Web mapping service. • Demonstrate maximum throughput for 1, 4, 8, 12, 16, 20, 24, 28, and 32 service instances on a 4-core

server.


Step 1: How can I demonstrate BATCH process service instance loads on the CPT Design tab?

Procedure: • Select the CPT Design tab. • Use copy sheet command to create new Design tab copy and label 4.1 (Lesson 4 question 1).

Best practice: It is good practice to move each new Design to a new tab and label newly created CPT tabs for future reference.

a) Use the Software Configuration module to configure software for a single platform configuration. • Open Software Configuration Module in CPT Design columns J through R.

CPT Design Software Configuration Module

• Software should be configured as shown.

o Select GIS as default platform tier in cells J5, L5, N5, and Q5. Note: You have just installed all workflow software on the GIS platform tier.

o All colored cells in workflow rows should be default. o Data Source for all workflows is SDE_DBMS.

• Close Software Configuration module and configure User Requirements.


b) Use the CPT Design Platform Selection module to select platform configuration. CPT Design Platform Selection Module

• Select following platform configuration for GIS: Platform Tier 09 in column B.

• Set GIS tier Fix Nodes to 1 in column H.

Note: You have just forced number of available servers = 1 for GIS tier.

c) Use Requirements Analysis module setup for batch workflow loads. • Select and delete any existing users or TPH throughputs (Requirements Analysis white cells) in columns C

and D. • Select following workflow in cell C6:

• Set peak concurrent users to 1 in cell C6. • Select minimum think time of zero “0” in cell AF6.

Note: BATCH process is created by selecting zero minimum think time. Warning: Entry in column F will override formula (cell turns white). Replace formula from another row if you need to reconfigure workflow for a different analysis.

• Set RESET function to ADJUST in cell AF2 and Blink to 10 in cell AG2.


CPT Design Requirements Analysis Module

1. Use the CPT Design model to complete the following table Service Instances

(cell C6) Throughput (TPH)

(cell H6) Response Time (sec)

(cell AH6) Server Utilization

(cells AH51:AH52) 1 9,118 0.39 24.9% 2 18,237 0.39 49.7% 3 27,334 0.40 74.5% 4 35,993 0.40 98.2% 5 36,629 0.49 99.9% 6 36,648 0.59 99.9%

Procedure: • Set RESET function to ADJUST in cell AF2 and Blink to 10 in cell AG2.

Note: Blink setting will need to be reduced to a smaller value to compute productivity during high utilization loads. Best Practice: Blink setting should start at 10 and reduce by factor of 10 until reaching a valid productivity.

• Enter number of Service Instances in cell C6 and Excel will complete the analysis.

Note: Cell C6 background turns GREEN when workflow productivity (E6) is valid. o Workflow throughput (TPH) is reported in cell H6. o Workflow response time (sec) is reported in cell AH6. o Server utilization is reported in cells AH51:AH52. Note: Configuration represents maximum batch process throughput with selected number of service instances. Single batch process consumes one processor core. Warning: Do not leave RESET function (cell AF2) in the ADJUST setting (ADJUST function leaves Excel in a circular calculation mode). Move to SAVE setting to save solution before leaving tab.


CPT Design tab configuration

Plot of CPT Design results for variable batch service instances in single server configuration

Note: Very little gain in throughput beyond 1 service instance per core (4 total instances). N+1 configuration (5 instances) ensures maximum server throughput load. Peak load response times increase with increasing number of service instances.


Step 2: How can I demonstrate Web mapping service instance loads on the CPT Design tab? Procedure: • Use copy sheet command to create a copy of 4.1 final design, move to end and label 4.2.

a) Use Requirements Analysis module setup for Web Service workflow loads. • Set peak concurrent users to 1 in cell C6. • Enter minimum think time of 0.01 in cell AF6.

Note: Minimum think time greater than zero will change CPT workflow model to represent Web service random arrival times. Small think time (0.01 sec) will simulate pooled service instance assignment supporting random arrival of user map requests.

• Select Test mode in cell E1. Note: Test mode changes the CPT model to calculate workflow productivity based on selected minimum think time. RESET ADJUST function (AF2:AF3) must be used to calculate a valid productivity. Throughput cells in column C and D will turn GREEN when calculating a valid workflow.

CPT Design Requirements Analysis Module

2. Use the CPT Design model to complete the following table Service Instances




(cells AH51:AH52) 1 7,682 0.46 20.9% 4 22,831 0.62 62.3% 8 30,370 0.94 82.8%

12 32,828 1.31 89.5% 16 33,936 1.69 92.6% 20 34,555 2.07 94.2% 24 34,947 2.46 95.3% 28 35,217 2.85 96.0% 32 35,414 3.24 96.6%



Procedure: • Set RESET function to ADJUST in cell AF2 and Blink to 10 in cell AG2.

Note: Blink setting will need to be reduced to a smaller value to resolve productivity during high utilization loads.

• Enter number of Service Instances in cell C6 and Excel will complete the analysis.

Note: Cell C6 background turns GREEN when workflow productivity (E6) is valid.

o Workflow throughput (TPH) is reported in cell H6. o Workflow response time (sec) is reported in cell AH6. o Server utilization is reported in cells AH51:AH52. Note: Configuration represents maximum Web Service throughput with selected number of service instances. Optimum throughput is reached at 4-6 service instances per core (16 instances for 4 core server). Warning: Do not leave RESET function (cell AF2) in the ADJUST setting (ADJUST function leaves Excel in a circular calculation mode). Move to SAVE setting to save solution before leaving tab.

Plot of CPT Design results for variable Web service instances in single server configuration

Note: Very little gain in throughput beyond 4 service instances per core (16 total instances). Web service configuration with 4-6 instances per core assures maximum server throughput load. Peak load response times will increase with increasing number of service instances.



1. What is the recommended number of service instances per core when configuring public Web mapping services for optimum throughput?

a. 1-2 service instance per core. b. 2-4 service instance per core. c. 3-5 service instance per core. d. 5-7 service instance per core.

2. What is the recommended number of service instances per core when configuring a map caching service for optimum throughput?

a. 1-2 service instance per core. b. 2-4 service instance per core. c. 3-5 service instance per core. d. 5-7 service instance per core.

ESRI | 4BGIS Data Architecture 5-1

Lesson 5: GIS Data Architecture

Purpose: Use the CPT Platform Capacity Calculator to demonstrate the impact of data source format selection on system performance and scalability.

Scenario: The selected data source format can have a considerable impact on workflow functional limitations as well as server and network processing loads. Data source selection is a configuration parameter separate from baseline workflow performance target selection, and data source format performance adjustments are made for each workflow during final system configuration. The data source options are different for Imagery workflows, and CPT configuration options change based on the selected software technology pattern. Data source selection is a primary consideration when selecting each workflow Enterprise design configuration.


• Configure Platform Capacity Calculator with Web mapping workflow. • Extend Platform Capacity Calculator to evaluate 6 Web mapping data sources. • Configure Platform Capacity Calculator for an ArcGIS image service workflow. • Extend Platform Capacity Calculator to evaluate 7 imagery data sources. • Configure CPT Calculator with custom ArcGIS image service workflow. • Add imagery workflow to Project Workflow list on Workflow tab.


Step 1: How does data source selection impact Web mapping platform capacity? Procedure: • Use Platform Capacity Calculator to demonstrate platform capacity. • Use copy sheet command to create new Hardware tab copy and label 5.1-3 (Lesson 5 questions 1-3).

Best practice: It is good practice to move to end and label newly created CPT tabs for future reference.

a) Modify Platform Capacity Calculator to include total of 6 workflow rows in column A. Procedure: Expose Platform Capacity Calculator workflows located behind the Platform Capacity chart.

Tip: Use your mouse to move Platform Capacity chart to the right exposing the white workflow cells in column A.

CPT Platform Capacity Calculator (Hardware tab) workflow selection

Note: The Platform Capacity Calculator generates performance range from selected medium complexity workflows, and will show an error if you do not select a medium workflow.

• Select one of the workflow rows and copy row. • Select another row within range of workflows and insert copies cells creating a new row. • Create total of 6 rows to complete the required configuration. • Select AGS10 REST MSD R 100%Dyn Med 10x7 JPEG for all workflows in column A.

Warning: Do not change formulas in any of the colored cells. All data modifications are made by selecting dropdown menus in the white cells.

• Select Xeon X5687 4 core (1 chip) 3600 MHz for platform configuration in column A. • Select TPH capacity output in column C. • Select Physical Platform in column E. • Once workflow rows are configured, replace the Platform Capacity chart back to its original position below

the “Select Platform Configuration” cell.


b) Use the modified Platform Capacity Calculator to evaluate data source options. Procedure: • Select the 6 available data source options in column I.

CPT Platform Capacity Calculator (Hardware tab) data source evaluation

1. What is the light complexity capacity with an SDE_DBMS data source? a. 133,200 TPH b. 35,600 TPH c. 106,600 TPH d. 53,300 TPH

2. What is the light complexity capacity with a small file geodatabase data source? a. 133,200 TPH b. 35,600 TPH c. 106,600 TPH d. 53,300 TPH

3. What is the light complexity capacity with a large shape file data source? a. 133,200 TPH b. 35,600 TPH c. 106,600 TPH d. 53,300 TPH


Step 2: How does data source selection impact Web Image service platform capacity? Procedure: • Use copy sheet command to create new Hardware tab copy and label 5.2 (Lesson 5 Step 2).

a) Modify Platform Capacity Calculator to include total of 7 workflow rows in column A. Procedure:

• Expose Platform Capacity Calculator workflows located behind the Platform Capacity chart. Tip: Use your mouse to move Platform Capacity chart to the right exposing the white workflow cells in column A.

• Select one of the workflow rows and copy row. • Select another row within range of workflows and insert copies cells creating a new row. • Create total of 7 rows to complete the required configuration. • Select AGS10 Imagery MosaicDS Dynamic 100%Dyn Med 10x7 JPEG for all workflows in column A.

Warning: Do not change formulas in any of the colored cells. All data modifications are made by selecting dropdown menus in the white cells.

CPT Platform Capacity Calculator (Hardware tab) workflow selection

• Select Xeon X5687 4 core (1 chip) 3600 MHz for platform configuration in column A. • Select TPH capacity output in column C. • Select Physical Platform in column E. • Once workflow rows are configured, replace the Platform Capacity chart back to its original position below

the Platform Configuration cell.


b) Use the modified Platform Capacity Calculator to evaluate data source options. Procedure: • Select the 7 available data source options in column I (do not include SDE DBMS).

CPT Platform Capacity Calculator (Hardware tab) data source evaluation

4. What is the light complexity capacity with a TIFF uncompressed data source? a. 77,200 TPH b. 50,200 TPH c. 139,000 TPH d. 53,300 TPH

5. What is the light complexity capacity with a TIFFLZW data source? a. 77,200 TPH b. 50,200 TPH c. 139,000 TPH d. 53,300 TPH

6. What is the light complexity capacity with a MRSID data source? a. 77,200 TPH b. 50,200 TPH c. 139,000 TPH d. 53,300 TPH


Step 3: How do I use the CPT Calculator to generate custom Imagery workflow performance targets?

a) Use the CPT Calculator to configure the following workflow. Workflow recipe

Tip: Make an Imagery selection in all cells that show a RED background. If Calculator results become unstable, complete RED cell selections and select RESET in cell E3. Calculator should stabilize as you return to your Imagery workflow selection.

b) Select the following user requirements and data source. User workflow requirements and data source

c) Select the following hardware architecture and network settings. Selected hardware architecture and network bandwidth


d) Deploy using a physical server platform two tier configuration. • Selecting DBMS in cell A11 hosts the Mosaic Dataset on the DBMS server.

CPT Calculator Configuration

• Use copy sheet command to save copy of the Calculator on a new tab and label 5.3 (Lesson 5 Step 3).


8. Identify the number of required ArcGIS Server licensed core (Web/SOC tier). a. 1 core b. 2 core c. 4 core d. 8 core


9. What software is generating the DBMS server load? a. Geodatabase b. Mosaic Dataset c. DBMS d. File share

Step 4: How do I add the custom Imagery workflow to my Project Workflows on the CPT Workflow tab?

Procedure: • Select and copy the project WebMap workflow row and "insert copied cells" to add a new workflow. • Locate the selected CPT Calculator workflow at the bottom of the CPT Workflow tab.

Note: Software technology performance factors for the custom workflow must be configured on the CPT Calculator tab.

• Select and copy blue calculator workflow cells in Column A through M. • Select the new project workflow cell in Column A and Paste Special/Values to insert the new Calculator

generated workflow service times into the new project workflow. Best practice: It is good practice to copy the workflow recipe (column A) and Paste Special/Values into the workflow description column (located in column AJ) to document workflow source.

• Insert Nickname (WebImage_) as the new workflow prefix.


CPT Workflow Tab Project Workflows

10. What is the total baseline processing time (column N) for the WebImage workflow? a. 0.105 sec b. 0.011 sec c. 0.022 sec d. 0.200 sec



1. What is the primary role of the ArcSDE geodatabase? a. Enterprise GIS data storage repository. b. Optimum Imagery publishing data source. c. Optimum multi-user GIS spatial feature editing database. d. Optimum GIS basemap publishing data source.

2. What is the primary role of a distribution or publishing geodatabase? a. Standard method for publishing imagery accessed through a mosaic dataset. b. Simple spatial feature data source for publishing dynamic operational business layers. c. Multi-generation one-way incremental ArcSDE replication service. d. Provides a platform for sharing static spatial data resources.

3. What is the primary purpose of an ArcSDE history geodatabase? a. Maintains a record of spatial database changes for geospatial temporal analysis. b. Maintains a record of spatial geodatabase changes for rollback and disaster recovery. c. Provides a time stamped record for establishing standard repeatable workflow processes. d. Provides data for analyzing information relative to time and location.

4. Under what condition is the ArcGIS Server Image Extension license required? a. Publishing preprocessed imagery using ArcGIS Server Image Service. b. Accessing raw imagery data sources from ArcGIS Desktop workstation. c. Publishing raw imagery using mosaic dataset with ArcGIS Server Image service d. Creating a mosaic dataset from raw imagery data sources.

5. What is the preferred way to move data from one schema to another? a. Geodatabase one-way replication. b. Geodatabase Transition (ETL). c. Database Replication Services. d. Storage Replication Services.

6. What is the preferred way to support a read only distribution geodatabase? a. Geodatabase one-way replication. b. Geodatabase Transition (ETL). c. Database Replication Services. d. Storage Replication Services.

7. What is the enabling technology supporting a distributed geodatabase architecture? a. Higher bandwidth network connections. b. Geodatabase versioning schema. c. Higher capacity commodity server platforms. d. All of the above.

ESRI | 5BNetwork Communications 6-1

Lesson 6: Network Communications

Purpose: Use the CPT to evaluate network bandwidth requirements.

Scenario: Network traffic contention is the most common cause of distributed GIS performance problems. Traffic contention occurs when network traffic throughput demands exceed available network capacity. Enterprise GIS deployment often significantly increases traffic over limited WAN and Internet service connections. Network connection bandwidth must be increased to handle peak traffic loads.


• Use CPT Calculator to select Web mapping specifications (recipe) that satisfies user performance needs. • Use CPT Design to demonstrate network bandwidth impact on user display response times. • Use CPT Design to identify reduced productivity expectations due to bandwidth congestion. • Use CPT Design to represent business requirements for a distributed system design. • Use CPT Design to identify required network bandwidth based on business requirements.

Step 1: How do I use the CPT Calculator to evaluate relative Web mapping display performance for my remote shared low bandwidth sites?

• Use copy sheet command to save copy of the Calculator on a new tab and label 6.1 (Lesson 6 Step 1).

a) Use the CPT Calculator workflow source selection to select project workflows. • Select Workflow tab as the Workflow Source in cell range C30:D30. • Select WebMap project workflow in cell range E30:I30.

Workflow source selection

Tip: The Workflow Source selection (C30:D30) allows you to select established Project Workflows or Standard Workflows for review on the Calculator tab.


b) Select the following user requirements and data source. User workflow requirements and data source

Note: Calculator Software Technology Performance Factors section is disabled when Workflow tab is selected as the Workflow Source.

c) Select the following hardware architecture and network settings (white cells). Note: Peak remote site concurrent users are 10 and 5 for this exercise, latency for Remote site 1 is 10 milliseconds, and network bandwidth (column G) is 3 Mbps for remote site 1 and 1.5 Mbps for remote site 2.

Selected hardware architecture and network bandwidth

Warning: Update white cells only. Colored cells contain formulas used to complete the CPT

Calculator design analysis.


d) Evaluate display response times for the WebMap workflow? CPT Calculator Configuration

1. Identify average display response times for clients on the local, remote site 1, and remote site 2 networks respectively?

a. 0.35 sec, 1.86 sec, 3.26 sec b. 3.26 sec, 1.86 sec, 1.35 sec c. 0.18 sec, 0.60 sec, 0.90 sec d. 0.90 sec, 0.60 sec, 0.18 sec


e) Evaluate display response times for the WebLite workflows? • Select the WebLite workflow in cell range E30:I30.

CPT Calculator Configuration

2. Identify average display response times for clients on the local, remote site 1, and remote site 2 networks respectively?

a. 3.26 sec, 1.86 sec, 1.35 sec b. 1.35 sec, 1.86 sec, 3.26 sec c. 0.18 sec, 0.60 sec, 0.90 sec d. 0.90 sec, 0.60 sec, 0.18 sec


Note: Proper software design selection can make a difference in user productivity.

Step 2: How does network bandwidth contention impact user productivity? • Use copy sheet command to save copy of the Design tab on a new end tab and label 6.2 (Lesson 6 Step 2).

a) Use the Design tab to demonstrate impact of network bandwidth on user productivity. • Select the same three workflows with 1 user each for the LAN, WAN, and Internet clients.

o DeskEdit_AGD10 wkstn MXD 100%Dyn Med 10x7 Feature o CitrixView_AGD10 Citrix MXD R 100%Dyn Lite 10x7 JPEG o WebMap_AGS10 REST MSD R 100%Dyn Med 10x7 JPEG

• Delete remote site rows (GREEN rows) on WAN and Internet networks. • Select bandwidth for LAN (100 Mbps), WAN (10 Mbps), and Internet (3 Mbps). • Select Intel Core i3-2120 client desktop platform and Xeon X5687 4 core servers.



3. Why does display response time increase when network bandwidth is reduced? a. Network latency decreases with reduced network bandwidth. b. Network transport time increases with reduced network bandwidth. c. Network latency increases with reduced network bandwidth. d. Network transport time decreases with reduced network bandwidth.

4. What is the projected data center Internet traffic? a. 1.667 Mbps b. 0.167 Mbps c. 0.200 Mbps d. 2.0 Mbps

5. What is WebMap Internet client workflow response time? a. 1.76 sec b. 1.05 sec c. 8.02 sec d. 9.5 sec


6. What is Internet WebMap network traffic flow? a. 1.667 Mbps b. 0.167 Mbps c. 0.200 Mbps d. 2.0 Mbps

7. What is the DeskEdit Internet client workflow traffic flow? a. 1.667 Mbps b. 0.167 Mbps c. 0.200 Mbps d. 2.0 Mbps

8. What is the DeskEdit Internet client calculated think time? a. 8.2 sec b. 5.0 sec c. -2.0 sec d. 0.28 sec

9. What does it mean for a workflow to have a negative calculated think time? a. Invalid workflow, need to increase user productivity. b. Invalid workflow, need to reduce user productivity. c. Valid workflow, need to reduce user productivity. d. Valid workflow, need to reduce user productivity.

b) Use the Design RESET function to reduce user productivity and identify valid solution. • Use copy sheet command to save copy of the 6.2 Design tab on a new end tab and label 6.2(2) (adjusted

solution tab for Lesson 6 Step 2). Procedure: • Set RESET function to ADJUST in cell AF2 and Blink to 10 in cell AG2.

Note: Cell C6 background turns GREEN and productivity cell return to normal when workflow productivity is valid.



Warning: Do not leave RESET function (cell AF2) in the ADJUST setting (ADJUST function leaves Excel

in a circular calculation mode). Move to SAVE setting to save solution before leaving tab.

10. What is the adjusted Internet DeskEdit workflow productivity? a. 20 DPM b. 10 DPM c. 6 DPM d. 6.66 DPM

11. What is the Internet DeskEdit workflow calculated think time? a. 4.3 sec b. 3.0 sec c. 5.2 sec d. 8.6 sec

12. What is the adjusted WebMap Internet client workflow display response time? a. 1.36 sec b. 1.72 sec c. 6.01 sec d. 0.50 sec


Step 3: How do I configure my Design to represent business requirements for local and remote site users?

• Use copy sheet command to save copy of the Design1 tab used in Lesson 2 on a new end tab and label 6.13 (Lesson 6 question 13). Note: In Lesson 2 our analysis was limited to evaluating the central data center network bandwidth requirements, and we did not evaluate bandwidth requirements for remote site locations. In this lesson we will add the remote locations to this design.

Copy of CPT Design for Lesson 2 question 12.

a) Identify peak number of users and WAN network connection bandwidth for each remote site location. During your business needs assessment in Lesson 2 you compiled the following information on user location and peak usage for each project workflow. The following chart identifies remote site user locations and network bandwidth connections for users located on the WAN. Expanded user needs summary


b) Expand CPT Design to include remote site locations. • Select and copy GREEN remote site row template located just below the Requirements Analysis module. • Select row just below the grey WAN row and use Excel insert copied cells function to insert GREEN remote

site row just under the Data Center WAN row. Tip: The GREEN remote site row template is configured to provide a default name for the remote site.

You can change the default name to identify each remote site as needed for your design.

• Select and copy GREEN remote site template before inserting each additional remote location. • Select row between the remaining WAN workflows and use Excel insert copied cells function to insert

second remote site. Warning: Always select a workflow row when adding remote sites. (Automatic naming convention will

not work properly when adding remote site immediately above a network row).

c) Follow procedures above to add one additional remote site network just below the Data Center Internet network GREY row.

Warning: Once you start including remote site locations, A GREEN remote site network row must be located just below each of the following Data Center GREY network rows.

• Enter remote site bandwidth connections in column H (1.5 Mbps for each remote site, 45 Mbps for the

public internet connections). CPT Design with remote site locations


d) Configure workflows for each remote site location. • Add an additional workflow for each remote site (each site will need a CitrixView and WebMap workflow).

Note: Workflows are added using same type of procedure used for adding remote site locations. Copy existing workflow row to use as a template and use the Excel insert copied cells function to add additional workflows. Tip: WebMap workflow row can be used as a template for adding a WebMap workflow to Remote site 1. CitrixView workflow row can be used as a template for adding a CitrixView workflow to Remote site 2.

• Select appropriate workflows for each site: • LAN clients (DeskEdit, DeskView, and WebMap) • Remote site 1 clients (CitrixView, WebMap) • Remote site 2 clients (CitrixView, WebMap) • Internet/Remote site 3 clients (WebLite) • Update peak user requirements in column C for each remote location workflow.

e) Complete configuration for network suitability analysis. • Network summation range must be adjusted for each remote site network in columns E and F. Column E of

each network identifies the total users and column F identifies the total network traffic for that network connection.

Tip: When updating a summation range, you can select the Excel formula row and it will show a blue box outlining the current summation cell range. You can select one of the corners of the blue box to extend the summation range. You should extend the network summation ranges (located in column E and F) to include the following network as shown in the figure above (G10:G13). Once configured, you can add or subtract workflows within a network range and the range will adjust to include the new workflows.


f) Identify required network upgrades. Once you configure the user workflows, update peak user requirements and update the network traffic summation ranges, Excel will complete the network suitability analysis. Invalid workflow productivity and network traffic bottlenecks will be identified by RED cells. In most cases you can remove workflow productivity issues by upgrading network bandwidth.

Best Practice: Use copy sheet command to save copy of the 6.3 Design tab on a new end tab for network upgrade purposes. Saving a sheet showing the network traffic bottlenecks is helpful when reviewing your design.

• As a rule of thumb, recommended network bandwidth is roughly twice the projected traffic loads. Tip. If network segments are rented from a service provider, these costs are normally monthly charges based on available bandwidth and are part of the Network Administrator’s budget. For this reason, bandwidth upgrades must be coordinated with the network administrator.

• For this design, make the following network bandwidth upgrades in column H: o Remote site 1 = 24 Mbps o Remote site 2 = 18 Mbps o Internet/Remote site 3 = 24 Mbps

CPT Design User Requirements Analysis module

Note. Network traffic column turns yellow when traffic exceeds 50 percent of selected network bandwidth. Selections close to 50 percent are often acceptable (you can review relative performance impacts on Workflow Performance Summary chart). Tip. Network bandwidth selections are based on standard offerings (T-1 = 1.5 Mbps, 2x T-1 = 3.0 Mbps, T-2 = 6 Mbps, 2x T-2 = 12 Mbps, 3x T-2 = 18 Mbps, 4x T-2 = 24 Mbps, T-3 = 45 Mbps, etc). Cell H3 allows you to input a unique bandwidth located at the top of the network bandwidth selection list.


g) Once you complete your network design, Excel completes the system architecture design analysis. CPT Design Enterprise solution

Note: The custom WebLite Internet traffic is much less than the WebMap services used in the earlier Lesson 2 design. Total ArcGIS Server processing load was also significantly reduced.

13. What is the Remote Site 1 network utilization? Note: Traffic cell turns YELLOW when traffic exceeds 50 percent of available bandwidth.

a. 46 percent b. 54 percent c. 43 percent d. 49 percent

14. What is the Internet/Remote site 3 network utilization? a. 46 percent b. 54 percent c. 43 percent d. 49 percent

15. What is the WAN/Remote site 2 network utilization? a. 46 percent b. 54 percent c. 43 percent d. 49 percent


16. What is the Internet WebLite average display response time? a. 0.37 sec b. 0.52 sec c. 0.54 sec d. 0.25 sec

17. What are the ArcGIS Server licensing requirements? a. Up to 2 core b. Up to 4 core c. Up to 6 core d. Up to 8 core

Note: ArcGIS Server licensing requirements were 8 core for the WebMap workflow in Lesson 2.



1. What is the most popular local area network protocol? a. ISDN b. HTTP c. Ethernet d. ATM

2. What is the planning factor used to convert MB of GIS Data to Mb of Network Traffic? a. 8 b. 10 c. 25 d. 100

3. How much GIS Data is needed to build a typical map display? a. 100 KB b. 10 MB c. 1 MB d. 1 GB

4. Identify data transfer time for 200 KB image display over a 10 Mbps network? a. 20 sec b. 2 sec c. 0.2 sec d. 0.02 sec

5. How does network latency impact display performance? a. Improves display performance by smoothing data transfer b. Primary concern when supporting Web clients c. Increases display response time for chatty communication connections d. Primary concern when supporting Windows Terminal clients

6. When does the network traffic cells turn RED? a. When network traffic exceeds available bandwidth. b. When response time exceeds network transfer time. c. When platform service time exceeds network transfer time. d. When traffic exceeds 50% of available bandwidth.

7. When does the workflow productivity cell turn RED? a. When minimum think time is zero b. When calculated think time is less than zero c. When "TEST" is selected in cell E1 d. All of the above

ESRI | 6BGIS Product Architecture 7-1

Lesson 7: GIS Product Architecture

Purpose: Use the CPT to identify the optimum GIS platform architecture.

Scenario: ArcGIS for Desktop can be installed on user workstations or centralized Citrix terminal servers. Web servers are configured in single-tier, two-tier, and three-tier architecture patterns. These configuration patterns can provide high availability, increase level of security, and deliver higher capacity systems. Hardware must be selected with the right capacity to deploy the required architecture pattern.


• Configure CPT Calculator tab to complete analysis for 3 tier WebMap services architecture. • Configure CPT Calculator tab to complete analysis for 2 tier WebMap services architecture. • Configure CPT Calculator tab to complete analysis for 2 tier WebMap virtual server architecture. • Configure CPT Design tab to complete analysis for 3 tier WebMap Enterprise architecture. • Configure CPT Design tab to complete analysis for 2 tier WebMap Enterprise architecture. • Configure CPT Design tab to complete analysis for WebMap virtual server architecture.


Step 1: How can I use the CPT Calculator to identify the optimum ArcGIS Server platform configuration?

• Use copy sheet command to save copy of the Lesson 6.1 Calculator tab on a new end tab and label 7.1-2 (Lesson 7 questions 1-2).

• Select WebMap workflow from the Workflow tab in row 30. • User requirements (75,200 TPH), SDE DBMS data source, Xeon X5687 4 core platforms.

a) Review platform requirements for a three tier physical server architecture. Procedure

• Select 3 Tier minimum platform architecture in cell A9.

1. Identify the minimum number of physical X5687 4 core platforms required to support this configuration.

a. 2 platforms b. 3 platforms c. 4 platforms d. 6 platforms


2. What is the ArcGIS Server licensing requirement? a. 2 core b. 4 core c. 6 core d. 8 core

b) Review platform requirements for a two tier physical server architecture. Procedure

• Select 2 Tier minimum platform architecture in cell A9. • Select WebMap workflow from the Workflow tab in row 30.

3. Identify the number of physical platforms required to support this configuration. a. 2 platforms b. 3 platforms c. 4 platforms d. 6 platforms



c) Review platform requirements for a two tier virtual server architecture (2 core servers). Procedure

• Select 2 Tier minimum platform architecture in cell A9. • Select VMware 2 core/node in column K for the server and DBMS tier.

5. Identify the number of physical platforms required to support this configuration. a. 2 platforms b. 3 platforms c. 4 platforms d. 6 platforms



Step 2: How can I adjust the CPT Design to show an Enterprise design solution with a three tier ArcGIS Server architecture?

• Use copy sheet command to save copy of the Design1 tab on a new end tab and label 7.7 (Lesson 7 Question 7).

User requirements and network settings from Lesson 2.12 Design tab.

a) Establish platform configuration for three tier Web architecture plus Windows Terminal Server tier. • Open platform tier 07 through 10 for configuring data center server environment in platform configuration

module. • Select Xeon X5687 4 core servers for all four platform tiers and Intel Core i3-2120 for client workstation in

column B. CPT Design Platform Configuration Module data center configuration

Note: Nicknames for each platform tier are followed by colon in column B. These nicknames will be used for selecting servers in the software configuration module.


b) Check requirements analysis configuration and complete software configuration. • WebMap services will be used for all Web workflows. • Peak data center workflow throughput requirements are same as provided in Lesson 2 and 6. • Open columns J through R to show Software Configuration Module.

Note: Columns E through H are hid in the graphic below for display purposes. CPT Design Software Configuration Module

• For this install, all workflow software components (colored cells in columns J through O and Q) should be

selected in default configuration (all software will be installed on default platforms in row 5). • Install Citrix software on WTS platform tier (cell K5). • Install Web software on Web platform tier (cell L5). • Install SOC software on GIS platform tier (cell N5). • Install DBMS software on DBMS platform tier (cell Q5). • Data source for all workflows (column R) is SDE_DBMS.


c) Review Enterprise design requirements for 3-tier ArcGIS Server configuration. • Close Software Configuration Module and review final design solution.

CPT Design Enterprise solution for 3-tier ArcGIS Server configuration.

7. How many Xeon X5687 4 core physical servers are required to support this solution? a. 2 platforms b. 3 platforms c. 4 platforms d. 6 platforms

8. What is the ArcGIS Server peak throughput capacity (below GIS platform tier in column G) for this configuration?

a. 130,000 TPH b. 1,066,000 TPH c. 171,000 TPH d. 485,000 TPH

9. What is the projected GIS server platform utilization? (44.1 percent) a. 46.3 percent b. 44.1 percent c. 31.0 percent d. 7.1 percent


Step 3: How can I adjust the CPT Design to show an Enterprise design solution with a two tier ArcGIS Server architecture?

• Use copy sheet command to save copy of the Lesson 7.11-13 Design tab on a new end tab and label 7.10 (Lesson 7 question 10).

a) Establish platform configuration for two tier Web architecture plus Windows Terminal Server tier. • Change the nickname for platform tier 09 to WebMap (this is where we will install the Web and Map server

software). CPT Design Platform Selection tier WebMap nickname assignment

10. Why is there no processing load on the new WebMap platform tier? a. Software moved to the Web server. b. System reconfigured to a two tier server architecture. c. No software installed on this tier. d. None of the above.

b) Configure two tier ArcGIS Server architecture • Use software configuration module to install Web and SOC software components on new WebMap server.


c) Review Enterprise design requirements for a two tier ArcGIS Server configuration. • Close Software Configuration Module and review final design solution.

CPT Design Enterprise solution for two tier ArcGIS Server configuration.

Note: You can close the Web tier (use group 1 row settings) since it is no longer supporting the configuration loads.

11. How many Xeon X5687 4 core physical servers are required to support this solution? (5 servers) a. 2 platforms b. 3 platforms c. 5 platforms d. 6 platforms

12. What is the ArcGIS Server peak throughput capacity (below WebMap platform tier in column G) for this configuration?

a. 130,000 TPH b. 158,000 TPH c. 485,000 TPH d. 978,000 TPH



Step 4: Can I reduce my ArcGIS Server license cost by using a virtual server environment? • Use copy sheet command to save copy of the Lesson 7 Step 3 Calculator tab on a new end tab and label 7.4

(Lesson 7 Step 4).

a) Change the WebMap platform tier from physical to virtual server configuration. • Select VMware in WebMap platform tier column I. • Select 2 core/node in WebMap platform tier column I.

b) Review Enterprise design requirements for ArcGIS Server virtual server configuration. • Once you change the virtual server configuration, the CPT Design tab will generate the final design solution.

CPT Design Enterprise solution for ArcGIS Server virtual server configuration.

14. What is the ArcGIS Server peak throughput capacity (below WebMap platform tier in column G) for this configuration?

a. 130,000 TPH b. 158,000 TPH c. 99,000 TPH d. 485,000 TPH




1. What is nota. ArcSDE GSRVR executables.

required on the database server to support geodatabase operations?

b. ArcSDE Schema. c. User Data Schema. d. ArcSDE Software License.

2. What must be installed on ArcGIS for Desktop to support an Oracle geodatabase direct connect architecture?

a. ArcSDE b. Oracle DBMS Client c. ArcSDE License File d. Geodatabase schema

3. What is the recommended ArcGIS for Desktop enterprise geodatabase connection architecture? a. ArcGIS Server Basic b. Application Server Connect c. Client Direct Connect d. File share disk mount

4. What is the most popular server based GIS power user application environment? a. ArcGIS for Server Map Services b. ArcGIS for Server Mobile ADF c. ArcGIS for Desktop deployed on Citrix Windows Terminal Server d. ArcIMS Image Service

5. ArcGIS 10 for Server Web services are executed on which server platform? a. Container machine b. Data Server c. Application Server d. Web Server

6. Which data source can be supported on a single-tier high available web server configuration? a. Single SQL Server geodatabase located on a SAN b. Single file geodatabase located on a NAS c. Mirror copy of SQL Server geodatabase located on each Web server d. Shape file data source supported on a file server

7. When is it ok to support ArcGIS for Server Web Applications over a Firewall? a. When needed to provide optimum enterprise security b. When using HTTP protocols to support enterprise applications in an SOA configuration c. When needed to support high availability requirements d. When needed to minimize network traffic across the firewall

ESRI | 7BPlatform Performance 8-1

Lesson 8: Platform Performance

Purpose: Use the CPT to evaluate candidate hardware platform performance and scalability.

Scenario: Hardware platform performance has doubled every two years while server cost keeps coming down, making a cost effective life cycle of server technology about 3-5 years max. Virtualization has improved IT flexibility in migrating servers from slower to faster higher capacity host platform environments – selecting the right server remains one of the most critical decisions in maintaining productive operations.


• Use the CPT Calculator to evaluate ArcGIS deployment on 2008 and 2011 4 core physical servers. • Use the CPT Calculator to evaluate ArcGIS virtual server deployment on 2011 8 core, 12 core, and 40 core

servers. • Use the CPT Design to evaluate ArcGIS deployment on 2008 and 2011 4 core physical servers. • Use the CPT Design to evaluate ArcGIS virtual server deployment on 2011 8 core, 12 core, and 40 core

servers. • Use the CPT Calculator to evaluate ArcGIS AMI deployment on the Amazon Cloud.

Step 1: Use the CPT Calculator to identify how upgrading to current technology can reduce overall operating cost?

• Use copy sheet command to save copy of the Lesson 7.5-6 Design tab on a new end tab and label 8.1-2 (Lesson 8 questions 1-2).

a) Evaluate projected WebMap loads on current 2008 hardware (X5260 4 core servers) Procedure

• Configure CPT Calculator with peak WebMap loads of 75,200 TPH, SDE DBMS data source. • Configure minimum two tiers direct connect platform architecture. • Select Xeon X5260 4 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Evaluate projected physical server loads.


CPT Calculator WebMap loads on X5260 4 core server.

1. What is the projected WebMap server utilization (cell H10) and ArcGIS for Server core licensing requirement (Cell T10)?

a. 69.7 percent, 3 core b. 47.6 percent, 2 core c. 69.7 percent, 12 core d. 47.6 percent, 8 core

2. What is the projected average local display performance (cell K12)? a. 0.86 sec b. 2.37 sec c. 0.35 sec d. 1.86 sec


b) Evaluate projected WebMap loads on 2011 hardware (X5687 4 core servers) Procedure

• Select Xeon X5687 4 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Calculator WebMap loads on X5687 4 core server.

3. What is the projected WebMap server utilization and ArcGIS Server core licensing requirement? a. 69.7 percent, 3 core b. 47.6 percent, 2 core c. 69.7 percent, 12 core d. 47.6 percent, 8 core


Step 2: Use the CPT Calculator to evaluate performance and cost of available 2011 virtual server configurations?

• Use copy sheet command to save copy of the Lesson 8 Step 1 Calculator tab on a new end tab and label 8.4 (Lesson 8 question 4).

a) Evaluate 2 core virtual serve performance on 2011 Xeon X5687 8 core host platforms. Procedure

• Select Xeon X5687 8 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for Web/SOC tier in column K. • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Calculator WebMap X5687 8 core server.

4. What is the projected WebMap server utilization and ArcGIS for Server core licensing requirement? a. 69.7 percent, 3 core b. 68.4 percent, 8 core c. 76.4 percent, 6 core d. 47.6 percent, 8 core


b) Evaluate 2 core virtual serve performance on 2011 Xeon X5690 12 core host platforms. Procedure

• Select Xeon X5690 12 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for Web/SOC tier in column K. • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Calculator WebMap E5690 12 core server.


6. What is the projected average local display performance? a. 3.62 sec b. 2.21 sec c. 0.70 sec d. 0.72 sec


c) Evaluate 2 core virtual server performance on 2011 Xeon E7-4850 40 core host platforms. Procedure

• Select Xeon E7-4850 40 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for Web/SOC tier in column K. • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Calculator WebMap E7-4850 40 core server.

7. What is the projected WebMap server utilization and ArcGIS Server core licensing requirement? (68.9 percent, 12 core)

a. 3.2 percent, 4 core b. 68.9 percent, 12 core c. 80 percent, 40 core d. 47.6 percent, 8 core

8. What is the projected average local display performance? (1.04 sec) a. 0.54 sec b. 1.04 sec c. 2.55 sec d. 3.95 sec


Step 3: Use the CPT Design to identify how upgrading to current technology can reduce overall Enterprise GIS operating cost?

• Use copy sheet command to save copy of the Design1 tab on a new end tab and label 8.9-10 (Lesson 8 questions 9-10).

• Configure as two tier WebMap architecture.

a) Evaluate projected Enterprise loads on current 2008 hardware (X5260 4 core servers). • Configure following projected user requirements and network bandwidth from Lesson 8 Step 3.

CPT Design Requirements Analysis module • Select Xeon X5260 4 core physical servers and i3-2120 client workstation in column B of the CPT Design

platform selection module. CPT Design Enterprise X5260 4 core server Enterprise GIS solution

9. What is the projected WebMap server utilization and ArcGIS Server core licensing requirement? a. 69.7 percent, 12 core b. 67.8 percent, 24 core c. 68.1 percent, 4 core d. 62 percent, 2 core

10. What is the projected average local WebMap display performance (cell AH8)? a. 0.87 sec b. 0.94 sec c. 0.96 sec d. 1.00 sec


b) Evaluate projected Enterprise loads on current 2011 hardware (X5687 4 core servers). • Select Xeon X5687 4 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Design Enterprise X5687 4 core server projected solution.


12. What is the projected average local WebMap display performance? a. 0.78 sec b. 1.14 sec c. 0.72 sec d. 0.34 sec


Step 4: Use the CPT Design tab to evaluate Enterprise performance and cost of 2011 virtual server configurations?

• Use copy sheet command to save copy of the Lesson 8.11-12 Design tab on a new end tab and label 8.13 (Lesson 8 question 13).

a) Evaluate 2 core virtual servers performance on 2011 Xeon X5687 8 core host platforms. Procedure

• Select Xeon X5687 8 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for all server tier in column I. • Evaluate projected physical server loads and ArcGIS Server licensing.




b) Evaluate 2 core virtual servers performance on 2011 Xeon X5690 12 core host platforms. Procedure

• Select Xeon X5690 12 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for all server tier in column I. • Evaluate projected physical server loads and ArcGIS Server licensing.





c) Evaluate 2 core virtual servers performance on 2011 Xeon E7-4850 40 core host platforms. Procedure

• Select Xeon E7-4850 40 core hardware platforms with 80 percent rollover (i3-2120 workstation). • Select 2 core/node VMware server configurations for all server tier in column I. • Evaluate projected physical server loads and ArcGIS Server licensing.

CPT Design Enterprise E7-4850 40 core server projected solution

Note: DBMS virtual server must be increased to 4 core to handle the peak loads.





1. What is an Arc10 performance baseline? a. ArcInfo 10 map display processing time b. Intel Xeon X5677 4 core 1 chip Server c. Platform with published SPECrate_int2006 per core baseline of 35 d. CY2010 most popular ArcGIS Desktop Platform

2. What is the preferred measure for Web platform capacity? a. Peak displays per minute b. Peak displays per hour c. Peak concurrent users d. Peak displays per minute per user

3. How do we measure platform performance? a. Vendor SPECrate_int2006 benchmark published baseline b. Internal Platform Performance Validation Test Results c. Project and customer feedback d. Vendor platform announcements

4. What technology contributed most to user productivity gains over the past 10 years? a. Esri software performance improvements b. Hardware platform performance improvements c. Customer performance expectations d. Expanding GIS marketplace

ESRI | 8BInformation Security 9-1

Lesson 9: Information Security Firewall Configurations

Identify the configuration:

1. Most secure firewall configuration (security manager’s perspective)? a. 1 b. 2 c. 3 d. 4

2. ArcGIS Server components in DMZ with data source on the private network. a. 1 b. 2 c. 3 d. 4

3. ArcGIS Server configured as a service supporting external Web applications across a firewall. a. 1 b. 2 c. 3 d. 4

4. All Web service components maintained and supported on the internal network. a. 1 b. 2 c. 3 d. 4

5. What are the three CIA tenets? a. Confidence, Intelligence, Authority b. Centricity, Intuitively, Aggregation c. Confidentiality, Integrity, Availability d. Central, Internet, Aggregate

SS

DS

WA

4

Private DMZ

SS

DS

WA

1

Private DMZ

Reverse-Proxy

SSDS

WA

3

Private DMZ

WA

2

Private DMZ

SS

DS

WA

Reverse-Proxy

optional

DBMS

ESRI | 8BInformation Security 9-2

6. What are the three types of security control? a. Internet, DMZ, Intranet b. Physical, Administrative, Technical c. Web, application, data d. Client, server, storage

7. What are the four areas of technical controls? a. Install, Deploy, Update, Maintain b. Plan, Test, Validate, Deploy c. LAN, WAN, Internet, Cloud d. Application, Host/Device, Network, Data

8. What data architecture provides the most security control options? a. Client/Server b. Web Applications c. Web Services d. Replication Services

9. What is the preferred ArcGIS Server firewall configuration? a. Web Service with Proxy Server b. Web Applications in DMZ c. All Web Services components in DMZ d. All Web Services components in DMZ except data source

ESRI | 9BPerformance Fundamentals 10-1

Lesson 10: Performance Fundamentals

Purpose: Use the CPT Test tab to translate benchmark test results to custom workflow service times that you can use to complete your system design. This exercise will also demonstrate compatibility between CPT Calculator, Test, and Design analysis capacity planning calculations.

Scenario: Rapid Web service development and the availability of simple Web service test tools is increasing the opportunity to use test measurements to generate custom workflow performance targets for use in planning. The CPT Test tab was developed to translate measured test results to workflow service times that you can use to complete your System Architecture Design. This exercise will use simulated test results to create custom workflows that you will use to complete your system design.


• Use the CPT Calculator to generate simulated test results. • Use the CPT Test tab to translate measured test results to custom workflow performance targets. • Use the CPT Design tab to validate custom workflow performance targets. • Use the validated test workflow to complete a CPT Design using new 2011 servers.


Step 1: How can I convert measured test results to generate a custom project workflow? The CPT Calculator can be used to generate server utilization from a given load test. We will use the CPT Calculator throughput input and generated output as our measured test results.

a) Configure the CPT Calculator with the WebMap workflow recipe. Procedure

• Use the Calculator tab to configure the following workflow.

• User Requirements (throughput input): 18,000 TPH • Platform Architecture: 3 tier, Minimum, DC • Hardware Platform Selection: Xeon 5160 4 core (2 chip) 3000 MHz

Note: We will use an older 2004 platform to demonstrate that you can use older platform technology for benchmark performance testing. You should make sure to use a production software configuration for best performance match.

• Client desktop: Intel Core i3-2120 2 core (1 chip) 3100 MHz CPT Calculator test workflow Xeon 5160 4 core platform

1. What is the throughput? a. 297,415 TPH b. 23,793 TPH c. 18,000 TPH d. 118,966 TPH


2. What is the measured utilization (Web,SOC,DBMS)? a. 4.8%, 60.5%, 12.1% b. 0%, 33%, 13% c. 0%, 65.3%, 12.1% d. 3%, 30%, 13%

b) Use CPT Throughput/Utilization Test tool to generate workflow service times. • Use the Test tab to enter the test results identified in Step 1a.

o Throughput (TPH) = 18,000 o Web application server (WAS) = 4.8% Utilization, 1 node, Xeon 5160 4 core server. o SOC machine = 60.5% Utilization, 1 node, Xeon 5160 4 core server. o DB machine = 12.1% utilization, 1 node, Xeon 5160 4 core server.

3. What are the Arc11 software component service times (Client, Web, SOC, DBMS)? a. 0.040 sec, 0.038 sec, 0.484 sec, 0.097 sec b. 0.040 sec, 0.048 sec, 0.605 sec, 0.121 sec c. 0.040 sec, 0.013 sec, 0.162 sec, 0.032 sec d. 0.040 sec, 0.026 sec, 0.324 sec, 0.064 sec

c) Move the test workflow results to your Project Workflows list. • Test results should be entered on CPT Test tab Throughput/Utilization tool in step 1b. • Create a workflow template by using Copy Row and insert copied cells below WebMap workflow in Project

Workflow section. TIP: Be sure to select row before using copy and insert commands.

• Select and copy blue cells in row below Test Workflows located at the bottom of the worksheet. • Select cell in column A of the new workflow template and paste special values to insert test workflow. • Label new workflow as “Test Workflow_WebMap Validation Test”.


CPT Workflow tab with new Test Workflow.

4. Does the test workflow results fall within the performance targets established for the WebMap workflow? (yes)

a. no b. yes

5. Where is the SDE service time for the test workflow? a. Web software includes SDE service times in a direct connect architecture. b. Client software includes SDE service times in a direct connect architecture. c. DBMS software includes SDE service times in a direct connect architecture. d. GIS Server software includes SDE service times in a direct connect architecture.


Step 2: How can I validate the new test workflow on the CPT Design tab? • Configure the CPT Design tab to represent the test configuration.

o Three tier platform architecture (Web, GIS, DBMS). o Select Xeon 5160 4 core server platforms and i3-2120 Client workstation in column B. o Select Test Workflow and insert 18,000 TPH in Requirements Analysis Module.

CPT Design test workflow with 5160 4 core servers.

6. What is the platform utilization profile (Web, GIS, DBMS)? a. 4.8%, 60.5%, 12.1% b. 0%, 33%, 13% c. 0%, 65.3%, 12.1% d. 3%, 30%, 13%

7. Does this utilization profile match with what we entered into the test tool? a. no b. yes


Step 3: How can I use the validated test workflow to generate capacity of my new 2011 servers?

• Use copy sheet command to save copy of the Lesson 10 Step 2 Design tab on a new end tab and label 10.3 (Lesson 10 Step 3).

• Select Xeon X5687 4 core server platforms and i3-2120 Client workstation in column B. CPT Design test workflow with X5687 4 core servers.

8. What is the peak GIS platform throughput for the 3 tier platform configuration (below GIS platform tier in column G)?

a. 1,075,000 TPH b. 427,000 TPH c. 85,000 TPH d. 1,422 TPH

9. What is the peak GIS platform throughput for a 2 tier platform configuration (reconfigure for 2 tier architecture and identify peak throughput?

a. 0 TPH b. 427,000 TPH c. 79,000 TPH d. 1,318 TPH



1. How do we address performance sizing? a. Establish performance targets based on user experience. b. Spend as much money as we can afford on hardware. c. Build simple applications. d. Buy the most current technology.

2. What is platform capacity? a. Service time + Queue time b. Transaction x service time c. Throughput x service time / # core processors d. Throughput / utilization

3. What is platform display service time? a. Total platform processing time to service a display request. b. Total time between client display request and resulting screen refresh. c. Individual platform client display response time. d. Total time for a platform to respond to a display request.

4. How can we calculate display service time? a. Display refresh time / #core processors b. Utilization x 60 x response time (sec) / (#core processors x throughput) c. 60 sec x #core processors / Platform Capacity (DPM) d. 60 x Display response time (sec) / platform capacity (DPM)

5. What causes queue time? a. Inbound transaction request rate exceeds peak processing capacity. b. Users think time exceeds system response time. c. Data transport times exceed system service time. d. System service times exceed user productivity rates.

6. What primary performance factor determines platform capacity? a. Platform response time b. Platform service time c. Platform queue time d. Platform utilization

7. What performance factor converts platform capacity from DPM to concurrent users? a. User response time represented in seconds per user display. b. Map service time represented in seconds per user display. c. Platform utilization in terms of peak concurrent users. d. User productivity represented by displays per minute per user.

ESRI | 10BCity of Portland 11-1

Lesson 11: City of Portland

Purpose: Use the CPT to complete a system architecture design case study.

Scenario: The following City of Portland user requirements represent a City Hall environment with three departments and two separate remote locations. ArcGIS Desktop is the client application used for spatial data maintenance and a geodatabase is used as the spatial database. ArcGIS Server is used to support Web Services. City of Portland User Needs

The City of Portland network architecture includes a central City Hall and two remote site locations (Maintenance Facility and Sandy River). Diagram of the network architecture is provided in the following graphic. Centralized Configuration Strategy


Remote sites are connected by a T-1 WAN, and public Web service access is provided through a T-1 internet service provider.


• Establish project workflows from recipes identified during the user needs assessment. • Use identified peak loads and CPT Design tab to complete year 1 network suitability analysis. • Provide recommended bandwidth upgrades for year 1 deployment. • Use CPT Design tab to complete identified hardware and software configuration. • Use CPT Design tab to identify year 1 hardware requirements. • Use identified peak loads and CPT Design tab to complete year 2 network suitability analysis. • Provide recommended bandwidth upgrades for year 2 deployment. • Use CPT Design tab to identify year 2 hardware requirements.

Step 1: What CPT workflows should I use to complete the City of Portland design? Use the capacity planning tool to complete the City of Portland requirements analysis. Provide a workflow to represent each of the ArcGIS software applications included in the user needs assessment. City Hall desktop clients will use existing workstations, remote desktop client will use Citrix Terminal clients, and Web Services will support Browser clients from the City Hall data center. The central geodatabase server will be located in the IT Data Center. Save a copy of the completed capacity planning workbook to document your design solution. Procedure

• Save copy of CPT as a new City of Portland workbook. • Establish Portland workflows on the CPT Workflow tab (assign workflow name)

ArcGIS Desktop Power users

(DeskPower_ArcGIS Desktop Power)

(CitrixPower_Remote ArcGIS Desktop Power)

ArcGIS Desktop Viewers

(DeskView_ArcGIS Desktop Viewer)

(CitrixView_Remote ArcGIS Desktop Viewer)

ArcGIS Server internal Web mapping application

(WebLocal_ArcGIS Server Internal Web Service)

ArcGIS Server REST MSD Public Web mapping application

(WebPublic_ArcGIS Server Public Web Service)

City of Portland Workflow Names and Description


City of Portland Workflow Performance Targets

1. What is the medium ArcGIS for Desktop client (DeskPower) Arc11 service time? a. 0.260 sec b. 0.380 sec c. 0.065 sec d. 0.040 sec

2. What is the ArcGIS for Server WebLocal Arc11 SOC service time? a. 0.260 sec b. 0.380 sec c. 0.065 sec d. 0.040 sec

3. What is the ArcGIS for Server WebPublic Arc11 SOC service time? (0.065) a. 0.260 sec b. 0.380 sec c. 0.065 sec d. 0.040 sec


Step 2: What is my projected network traffic for City Hall year 1 Design? Procedure: • Use copy sheet command to create a copy of CPT Design tab, move to end and label CHy1 (City Hall year 1). • Configure two remote sites on Data Center WAN and one remote site on Internet. • Configure user workflows for each network location.

o 3 workflows on LAN (13 DeskPower, 9 DeskView, 10,000 TPH WebLocal) o 3 workflows on Maintenance Facility (38 CitrixPower, 32 CitrixView, 4,000 TPH WebLocal) o 2 workflows on Sandy River (15 CitrixView, 1,000 TPH WebLocal) o 1 workflow on Public Internet (48,000 TPH WebPublic)

• Select required workflows in column B from Project Workflow list. • Enter peak concurrent users in column C and projected TPH in column D. • Enter available network bandwidth in column H.

o 1,000 Mbps for LAN o 1.5 Mbps for WAN, Internet, and remote locations.

• Update summation ranges for network traffic in column E and users in column E. CPT Design Network Suitability Analysis for City Hall Year 1.

Note. System performance bottlenecks are identified by the RED cells. Network traffic is more than the available bandwidth – will need to upgrade network to complete the design.


Step 3: What are the City Hall Year 1 network bandwidth requirements? Procedure: • Use copy sheet command to create a copy of the CHy1 tab, move to end and label CHy1(2). We will use

this new worksheet to identify network upgrade requirements. • Use network bandwidth selections in column H to select a bandwidth that is roughly twice the projected

traffic loads. Best Practice. Select closest bandwidth below 60 percent utilization for starting point. You can review

relative performance impacts on Workflow Performance Summary chart. City of Portland Year 1 recommended network bandwidth upgrades.

Note: Lower bandwidth connections are more sensitive to network congestion. For example, Sandy River network utilization is 51 percent with network traffic reducing WebLocal display performance by over 100 percent. Maintenance Facility network utilization is 58 percent with network traffic reducing WebLocal display performance by less than 25 percent. Workflow Performance Summary chart along with network upgrade costs can be used to determine optimum design solution.


Step 4: What is the recommended City Hall hardware configuration? IT department plans to deploy in a virtual server environment with a separate GIS Server platform to host WebLocal and WebPublic services. Web services will be deployed in a two tier architecture using a common DBMS data source.

Procedure: • Establish 4 tier architecture in platform selection module and update nicknames for each platform tier

(WTS, WebLocal, WebPublic, and DBMS). • Select host platform for each tier in column B.

o Intel Core i3-2120 2 core workstation will represent client desktops. o Xeon X5690 12 core (2 chip) 3466 MHz will support all virtual server host platforms.

Best Practice. Virtual server host platform selection may impact user productivity and software licensing. Select an optimum platform to satisfy overall business needs. Do not overprovision host server resources (include dedicated host platform physical core and memory for each virtual server configuration).

• Select VMware and 2 core/node in column I for each tier. Caution. Customers are reporting poor performance when deploying XenApp in a VMWare virtual server

environment. CPT analysis will assume normal virtual server performance. • Select 80 percent rollover for each tier in column H.

Platform Selection Module


Step 5: What is the recommended City Hall year 1 software configuration? Software will be configured to separate WebLocal and WebPublic GIS Server software environments.

Procedure: • Use copy sheet command to create a copy of the CHy (2) tab, move to end and label CHy1 (3). We will use

this new worksheet to complete the workflow software configuration. • The Software Configuration module in columns J through R can be opened by selecting the first column

group (+) located above column S. You can hide Columns E through I for simpler display of the software configuration.

• Select default software install platforms in row 5 (Client, WTS, WebLocal, Default, DBMS). WebLocal tier will be default install for Web and SOC software components. Note: SDE Default selection specifies direct connect architecture for all default workflows.

• Select WebPublic tier for Web and SOC software components for the ArcGIS Server Public Web Service workflow.

• Select SDE_DBMS data source for all workflows in column R.

Step 6: What are the hardware requirements for City Hall year 1? Once the CPT Requirements Analysis, Network Suitability, Platform Selection, and Software Configuration is complete, Excel will complete the system architecture design.

Procedure: • Use copy sheet command to create a copy of the CHy (3) tab, move to end and label CHy1 (4). We will use

this new worksheet to review System Architecture Design results for City Hall year 1. • Unhide all columns and select column group 1 to show the final design solution.


City Hall Year 1 System Architecture Design results

4. Identify the minimum WTS tier server configuration? a. Two 2-core virtual servers b. Two 12-core servers c. Five 2-core virtual servers d. One 2-core virtual servers

5. Identify the minimum WebLocal tier server configuration? a. Two 2-core virtual servers b. Two 12-core servers c. Five 2-core virtual servers d. One 2-core virtual servers

6. Identify the minimum WebPublic tier server configuration? a. Two 2-core virtual servers b. Two 12-core servers c. Five 2-core virtual servers d. One 2-core virtual servers


7. What is the total ArcGIS for Server licensing requirement? a. Up to 4 core b. Up to 6 core c. Up to 8 core d. Up to 10 core

8. How many 12 core physical host platforms support this configuration? a. One 12-core server b. Two 12-core servers c. Three 12-core servers d. Four 12-core servers

Step 7: What is my projected network traffic for City Hall year 2 Design? Procedure: • Use copy sheet command to create a copy of the final City Hall Year 1 CPT Design tab, move to end and

label CHy2 (City Hall year 2). • Update Peak Concurrent users and projected throughput for City Hall Year 2.

o 3 workflows on LAN (53 DeskPower, 17 DeskView, 20,000 TPH WebLocal) o 3 workflows on Maintenance Facility (108 CitrixPower, 40 CitrixView, 12,000 TPH WebLocal) o 2 workflows on Sandy River (15 CitrixView, 3,000 TPH WebLocal) o 1 workflow on Public Internet (96,000 TPH WebPublic)

• Network upgrades from Year 1 should still be in column H. o Data Center: 1,000 Mbps for LAN, 45 Mbps for WAN, and 24 Mbps for Internet

• Remote locations: 24 Mbps for Maintenance Facility, 6 Mbps for Sandy River, and 24 Mbps for Public. Note: General Public Internet Client bandwidth should be set same as the Data Center Internet.


CPT Design Network Suitability Analysis for City Hall Year 2.

Note. System performance bottlenecks are identified by the RED cells. Network traffic is more than the available bandwidth – will need to upgrade network to complete the design.

Best Practice. Recommended bandwidth upgrades are roughly twice projected peak traffic flow. Upgrades are normally multiples of T-1 (1.5 Mbps), T-2 (6 Mbps), or T-3 (45 Mbps) values.

9. What is the peak traffic for the Data Center WAN? a. 113.6 Mbps b. 35.5 Mbps c. 26.7 Mbps d. 31.3 Mbps

10. What is the peak traffic for the Maintenance Facility gateway? a. 113.6 Mbps b. 35.5 Mbps c. 26.7 Mbps d. 31.3 Mbps


Step 8: What are the hardware requirements for City Hall year 2? Procedure: • Use copy sheet command to create a copy of the CHy2 tab, move to end and label CHy2 (2). We will use

this new worksheet to review System Architecture Design results for City Hall year 1. • Use network bandwidth selections in column H to select a bandwidth that is roughly twice the projected

traffic loads. Best Practice. Select closest bandwidth below 60 percent utilization for starting point. You can review

relative performance impacts on Workflow Performance Summary chart. • Once the CPT Requirements Analysis, Network Suitability, Platform Selection, and Software Configuration

are complete, Excel will complete the system architecture design. Note. We can use the same platform selection as we did in year 1 for our initial design.

City Hall Year 2 System Architecture Design results

11. Identify the minimum WTS tier server configuration? a. Three 2-core virtual servers b. Four 12-core servers c. Ten 2-core virtual servers d. One 6-core virtual server


12. Identify the minimum WebLocal tier server configuration? a. Three 2-core virtual servers b. Four 12-core servers c. Ten 2-core virtual servers d. One 6-core virtual server

13. Identify the minimum WebPublic tier server configuration? a. Three 2-core virtual servers b. Four 12-core servers c. Ten 2-core virtual servers d. One 6-core virtual server

14. What is the total ArcGIS for Server licensing requirement? a. Up to 4 core b. Up to 8 core c. Up to 10 core d. Up to 12 core

15. How many virtual core are required to provide a single DBMS server? a. Three 2-core virtual servers b. Four 12-core servers c. Ten 2-core virtual servers d. One 6-core virtual server

16. How many 12 core physical host platforms support this configuration? a. Two 12-core server b. Three 12-core servers c. Four 12-core servers d. Five 12-core servers

ESRI | Answer Sheet

Exercise answers Question Lesson

1 Lesson

2 Lesson

3 Lesson

4 Lesson

5 Lesson

6 Lesson

7 Lesson

8 Lesson

9 Lesson

10 Lesson

11 1 d a b * c a c c d c b 2 b c c * a c d a c a a 3 c b d b b b d b c c 4 b a b c d d c a b c 5 a d c a a a b c d a 6 b a a b c c c b a a 7 c b b d a d b d b c 8 d c d c c c b c c b 9 a d b b b a a c b 10 c b d d c a d 11 b c b c d c 12 a b a b d a 13 c a b d a a 14 c c d c c d 15 a c c c d 16 b d b c 17 a b b

Review question answers Question Lesson

1 Lesson

2 Lesson

3 Lesson

4 Lesson

5 Lesson

6 Lesson

7 Lesson

8 Lesson

9 Lesson

10 Lesson

11 1 d a d c c c a c NA a NA 2 d b c a b b b b d 3 c c a a c c a a 4 a d c c c c b c 5 d c c b c a a 6 d d a a c b 7 c b d b d

* Lesson 4 Question 1 Service Instances




(cells AH51:AH52) 1 9,118 0.39 24.9% 2 18,237 0.39 49.7% 3 27,334 0.40 74.5% 4 35,993 0.40 98.2% 5 36,629 0.49 99.9% 6 36,648 0.59 99.9%

* Lesson 4 Question 2 Service Instances




(cells AH51:AH52)

1 7,682 0.46 20.9% 4 22,831 0.62 62.3% 8 30,370 0.94 82.8% 12 32,828 1.31 89.5% 16 33,936 1.69 92.6% 20 34,555 2.07 94.2% 24 34,947 2.46 95.3% 28 35,217 2.85 96.0% 32 35,414 3.24 96.6%

system architecture design exercises may 2012

Documents