• No results found

Introduction. About the authors


Academic year: 2021

Share "Introduction. About the authors"


Loading.... (view fulltext now)

Full text


Layer, Select, Focus (LSF) View: Visualization and

Perspective in Enterprise Networks

by Maria Velez-Rojas, Serge Mankovskii, Steven Greenspan, Research Staff

Members, CA Labs, Michael Roberts, Senior Product Designer, CA Technologies,

and Esin Kiris, Technology Architect, Syncsort


Managers and operators of modern Enterprise IT environments deal with massive amounts of information in a very stressful environment. They routinely manage hundreds of thousands of interconnected elements while collaborating with multiple business stakeholders under strict time constraints. This paper proposes a visualization method to simplify managing large

amounts of information called the LSF view method. The LSF view method simplifies visualization by generating overviews of the network emphasizing the connections that are most salient to the user. This method also provides users with three key capabilities: manipulating of layer semantics, selecting instances of relevant concepts, and focusing on relevant information.


Management of information technology (IT) infrastructures in the modern Enterprise is a challenge even for experienced IT professionals. IT managers and IT analysts need to monitor the performance of a wide range of physical and logical elements interconnected in a complex network of physical connections, logical dependencies, across multiple levels of abstraction and composition. At the same time, IT practitioners are required to meet certain key performance indicators (KPIs) such as minimum resources availability, mean time between failure and mean time to repair. Understanding the status of network elements and their dependencies is crucial for analyzing how problems propagate and for identifying the stakeholders that need be engaged to solve a problem as soon as a change in an element status is reported.

Visual representations of data are used in IT management solutions to communicate patterns and facilitate the exploration of the complex relations between elements. With data visualizations, users can identify patterns and trends that would be otherwise hidden on massive text-based reports and tables. It is this potential to communicate complex information at a glance that is making data visualizations pervasive in IT management environments. For instance, Dashboards often use charts (e.g. pie charts and histograms) to provide executive summaries of important metrics as shown in Figure 1. More complex data visualizations such as the node-link diagram in Figure 2 are often used to provide detailed technical information about network elements and their relationships.

Data storage, data management and visual interfaces have matured over the last decade and they have successfully been adopted in corporate

environments. However, the integration of data visualizations as part of IT management solutions has been modest. Although it’s technologically possible to create data visualizations capable of representing large-scale databases (e.g.,

1, 2), other important aspects need to be addressed for data visualizations to

become pervasive on IT management. Factors such as ineffective data representation, lack of data exploration capabilities and complex user interfaces have been found4to inhibit the adoption of data visualizations.

About the authors

Dr. Maria Velez-Rojas is a Research Staff Member with CA Labs, based in San Francisco, California. She joined CA Labs in 2010. Her current research focus is on the design of usable and efficient visualization techniques of complex Information Technology environments. In addition to her research in visualization, Maria is working in the evaluation of technologies for the efficient

processing of data base transactions. She is also directing human factors research aspects in projects related to usable security. Maria received a Master of Science and a PhD degree in Computer Engineering from Rutgers University in 2004 and 2009, respectively.

Serge Mankovski is a Research Staff Member with CA Labs. He is based in Redwood City, California. His current research interests include Big Data, Visualization of very large IT


In this paper we present a novel data visualization approach designed using the requirements identified in interviews with IT professionals regarding the role data visualization play on their workflow. We propose a data visualization that creates simplified overviews of complex IT environments by emphasizing areas of the network and connections that are relevant for the user. This visualization provides means to explore the data by selecting the semantics that define layers within the network environment, selecting instances of relevant elements or areas, and interactively filtering irrelevant information.


We interviewed 20 IT professionals with 2 to 15 years of experience from the areas of change and configuration, IT service delivery, software engineering,

infrastructures, and Predictive Analytics. Serge has been active in industry sponsored research for over nine years. He has 19 filed patent applications, and has been granted 11 granted in the United States , United Kingdom and Canada. Serge received a Master‘s degree in Automatic System Control from the Kishinev Polytechnic Institute in Moldova.

Dr. Steven Greenspan is a Vice President at CA Technologies and a Research Staff Member in CA Labs. Please view his biography in the article “The Human in the Center” in this issue, or at www.ca.com/calabs.

Michael Roberts is a user experience researcher and designer currently working at CA Technologies leading User Interface design for a large mainframe management product. He has evangelized user-centered design and driven user research and design for numerous networking, IT management, and telephony products. He holds an MA in Engineering Psychology from New Mexico State University. Michael lives in McKinney, Texas with his wife, three sons, two dogs, and more projects around the house than he can count.

Figure 1 (b). Dashboard designs with elements commonly used on IT management solutions. (a) By BPMonline (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/li-censes/by-sa/3.0) or GFDL. (b) By 4zHaR screenshot done by me (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/[CC-BY-SA-3.0)], via Wikimedia Commons (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons


infrastructure engineering and software architecture and as described on Table 1. The goal of our interviews was to understand how everyday work activities of IT professionals are supported by data visualizations and to identify the issues that arise when the complexity of the network infrastructure increases. We asked a series of open-ended questions to capture the workflow, current use of visualization tools and needs of the interviewees.

We divided the interview in three parts. First, we asked questions related to the participant’s work environment. In particular we asked the user to describe in detail:

Work activities: Personal or organizational goals and actions carried out to •

pursue these goals.

Context: user roles defined explicitly defiled by the company and implicitly •

developed as part of the job requirements.

Work artifacts: information retrieved, created and modified during work •


Part 2 of the interview focused on the data visualization used to work with this information. We asked participants to describe how they currently use data visualization tools to perform their jobs and what value they see in them. We also asked participants to think of any area of their job that would benefit from having a visualization tool available without limiting themselves to the

functionality currently provided by their IT management solutions. We concluded the interview by asking participants to reflect back on their answers and to think of any comments of information that they would like to share with us.

Challenges in Complex Environments

Participants were asked to identify the main challenges faced when working with very large network environments and to discuss the role that visualization tools play on those scenarios.

Participants reported that the single most important variable determining the value of an IT management solution is support to user’s tasks that results on minimization of the time it takes for IT professionals to perform their job. Two factors were identified as having a significant impact in the time it takes them to perform their work activities: Ease of navigation and exploration of the data related to the task at hand and collaboration and communication with multiple stakeholders. Although the details of the workflow for each IT professional were different, these two factors were prevalent among all interview participants.

Dr. Esin Kiris is a Technology Architect at Syncsort, currently leading the innovative data integration product strategy and design initiatives with responsibility to level the products to the emerging technologies, such as Big Data, Cloud and Virtualization. Prior to Syncsort, Esin held several positions at CA Technologies including Principal Product Design Architect, Senior Director and Senior User Interface Designer.

Esin received a Ph.D. and M.S. in Industrial Engineering from Texas Tech University, specialized in UI design, CHI and experimental design, and a B.S. in Computer Engineering from Bogazici University, Istanbul, Turkey.

Table 1. Profile of IT professionals interviewed Profile Description

Change and

configuration Ensure that the risk of making any change is understood by all partiesinvolved IT service

delivery Manage the needs and requests of customers in the shortest time possible Software

engineering Design products and interfaces based on user requirements Infrastructure

engineering Design and implement monitoring infrastructure, provide feedback on perceived solution design flaws and recommending product improvements


architecture Define requirements and design solutions, ensure that all products orimplementations follow common architectural guidelines

Participants reported that the single most important variable determining the value of an IT management solution is support to user’s tasks that results on minimization of the time it takes for IT professionals to perform their job.


Navigation and Exploration

Interview participants expressed their preference for interfaces designed to present an overview of their workspace (e.g., list of open customer tickets that require attention, list of alarms triggered by network components, etc.), that provides them with the “minimum detailed information“ required to related to take the next step on task to be performed. Node-link diagrams (NLDs) of network infrastructure (as the example shown on Figure 2) were identified as visualizations with scalability and navigability issues.

NLD visualizations can display millions of elements and provide tools to improve navigation such as intelligent zoom and network overviews2. However, overviews often do not reflect the way IT professionals think about their network

environments. Participants also noted that after zooming in, visual search of elements and properties on the screen is a complex task due to all the “noise” or unnecessary information.

IT Professionals compartmentalize the Enterprise according to the aspects they are interested in. For instance, service assurance managers divide the infrastructure according to abstract business services (e.g., payroll service, email service, etc.) and the physical elements that are part of them. At the same time, infrastructure engineers can see the Enterprise as a set of interconnected sub-networks that are distributed across multiple locations. Therefore, the overview of the

Enterprise infrastructure and relevant detailed information varies according to the task that the IT professional must perform.

Visual overviews provided by network infrastructure visualization solutions mostly focus on aggregating elements based on structural properties of the IT environment and the semantic areas of IT networks are represented as superimposed layers that contain certain types of elements on the infrastructure. Just like blueprints, these layers can be dynamically added or subtracted to the NDL visualization according to the needs of the user.


We designed and implemented a prototype for a visualization tool that presents a semantically rich overview of the network elements within a context relevant to the user. We allow users to define “layers” within the network

infrastructure that represent the areas of information they are interested in. As shown on Figure 3, each layer is defined by a subset of the elements in the network structure and a subset of the relations between elements.

Figure 2. Node-Link diagram visualizations used to represent the elements and relations in networks infrastructure (a) By Unforgettable fan en:User:Unforgettable fan [Public domain], via Wikimedia Commons. (b) By Luca Saiu, Jean-Vincent

Loddo/Marionnet. (www.marionnet.org) [GPL (http://www.gnu.org/licenses/gpl.html) or CC-BY-SA-3.0 (http://creativecommons.org/li-censes/by-sa/3.0)], via Wikimedia Commons


The elements on each layer are then viewed as Euler Diagrams. Euler diagrams are a way to intuitively depict the union and intersection of sets [5]. They consist of closed shapes (e.g., circles, rectangles or other polygons) that depict sets of elements. The overlap between shapes represents common or shared elements as shown on Figure 4.

This data visualization supports fast navigation of large environments and simple transition between overview and detailed data, while retaining the information displayed within a context relevant to the task being performed by the IT professional. It also provides the capability to see dense relations in a summarized form.


Users navigate and explore the infrastructure by selecting and un-selecting different regions of the layer. As shown on Figure 5, the overlaps of the layer regions correspond to relationships between sets of elements in the layer. The selection mechanism in the Layer view allows informal expression of formal logical query used as filter expressions for graph visualizations.

Multiple layers can be defined with a network environment and the relations of elements between layers can be

examined. Let’s consider the following case. An application server is affected by power supply; within server there is one operative system and the operative system hosts many applications that interact to each other. The propose visualization allows IT professionals to explore the relations between the elements at the extremes in this infrastructure and answer the following question: How are applications impacted by one single power supply?

Figure 4. Layer elements dependencies selection. A subset of the resources connected to the layer elements is selected based on their properties and type of relation. The elements on each layer are viewed as areas that overlap depending on the resources they share. Figure 3. Layer element selection. Elements are assigned to different layers within the network based on their properties.

(a) (b)


As illustrated on Figure 6, users can define four layers in the network environment: Power supply layer, application servers layer, operative system layer and applications layer. By selecting an element on the power supply level, all the

applications related to that particular power supply are clearly visible. In this case, database queries connecting are expressed intermediate layers (i.e., the application servers layer and operative system layer), and result of the queries are visible as dependencies between elements at the end layer (i.e., power supply layer and applications layer).

When the regions of interest are selected, users can analyze the detailed information in those areas by displaying the selected elements and relations on data visualizations that emphasize the sub-network’s structure, semantics (i.e., the mining of elements and relations properties), temporal or statistical properties.

We can also extend the semantics of this data abstraction beyond exploration to action. For example, when a region is clicked, the log and alert entries, and emails related to those assets could be displayed, work flows for notification could be staged, etc.

Graph representation creates a level of abstraction above the data sources. It is possible that data for one layer would come from a data source dramatically different from data in another layer. Moreover, within one layer it is possible that elements from the top set would come from a data source different from the source of bottom elements. The LSF model creates an opportunity for establishing visual relationships within data sources. Set semantics of Eurler diagrams enable creation of filter expressions in different languages.

Figure 5. Selection of areas on the layer and equivalent selection syntax

Figure 6. Selection of four network layers: (i) Power supply layer, (ii) application servers layer,

(iii) operative system layer and (iv) applications layer. Dependencies between elements on the different levels can be explored.


Collaboration and Communication

Management of large IT enterprise environments is a collaborative endeavor at multiple levels across the organization. Independent groups work simultaneously on different aspects of the network and they must have well-established communication channels in order to avoid conflicts.

Communication between groups involves weekly meetings and continuous asynchronous collaboration within and between business units. Some of the visualization tools provide simple annotation capabilities, but most are not well integrated with common collaboration tools such as email and messaging applications.

Interview participants require communication tools that will enable them to see activities taking place in multiple areas of the IT environments, to explore the history of activities and comments left by others and to share insights that will allow other to address issues in an effective way. IT professional see a great potential for visualization tools to be the communication artifact that establishes common ground among stakeholders, supports conversation, and sustains conservation and change. However, our interviews indicate that this requirement is often neglected in tool design.

Participants of IT management process possess a variety of skills and seek to accomplish different, sometimes conflicting goals. For example, operations personal strives to maintain infrastructure in operational state and reduce amount of change. Business development and asset management personnel evolve and update the infrastructure, reflecting changes in technology and business requirements. While the participants know details of their own domain of expertise, no one has total knowledge and understanding of all aspects of the management process in the enterprise. The IT management tools and their interface become primary source of shared understanding and means of achieving synergy among the participants enabling them to resolve problems and develop new capabilities.

We are investigating how to best incorporate communication models similar to those found on social networking sites such as Facebook to support the management of large IT networks. Through these tools we aim to capture the

perspectives that each stakeholder has with respect to the elements that are part of their individual workspace. For instance, by looking at the comments left by other stakeholders, a change engineer can select a date and time to program an update in one of the hardware components of the network. At the same time, by adding the scheduled change to the social exchanges related to this element, other stakeholders will be aware of the upcoming update thus avoiding potential conflicts.

Social exchanges can also provide insight to new IT staff on who the stakeholders for each network component are and what workflow senior staff members follow. As shown on Figure 7 comments by IT practitioner can be added within the context of an alarm, they could be related to IT metrics, to element properties such as it location or to the priority of the element.


Future Work

As part of our future work, we are investigating the improvement of the social network communication tools and

considering a few other collaboration features to be included in our visualization design such as rating model based on the exchanges made by practitioners. Informal presentations have been well received by customers and a quantitative study is necessary to determine the usability of these visualization tools.


J. Abello, F. van Ham, N. Krishna, “ASK-GraphView: A Large Scale Graph Visualization System”. In IEEE Transactions on 1

Visualization and Computer Graphics, Vol. 12, No. 5, September 2006.

I. Herman; G. Melancon; M.S. Marshall, “Graph visualization and navigation in information visualization: A survey”. In 2

IEEE Transactions on Visualization and Computer Graphics, Vol. 6, Issue 1, pages 24 – 43, January 2000.

S. Noel, S. Jajodia, “Managing attack graph complexity through visual hierarchical aggregation”. In Proceedings of the 3

2004 ACM workshop on Visualization and data mining for computer security, pages 109M118, 2004. P. Hanrahan, C. Stolte, J. Mackinlay, “Visual Analysis for Everyone”. Tableau White paper, January, 2007 4

Rosario De Chiara, Mikael Hammar, and Vittorio Scarano. 2005. A System for Virtual Directories Using Euler Diagrams. 5

Electron. Notes Theor. Comput. Sci. 134 (June 2005), 33-53.

George Robertson, Kim Cameron, Mary Czerwinski, and Daniel Robbins. 2002. Polyarchy visualization: visualizing 6

multiple intersecting hierarchies. In Proceedings of the SIGCHI conference on Human factors in computing systems: Changing our world, changing ourselves (CHI '02). ACM, New York, NY, USA.


Copyright © 2013 CA. All rights reserved. All trademarks, trade names, service marks and logos referenced herein belong to their respective companies.

The information in this publication could include typographical errors or technical inaccuracies, and CA, Inc. (“CA”) and the authors assume no responsibility for its accu-racy or completeness. The statements and opinions expressed in this publication are those of the authors and are not necessarily those of CA.

Certain information in this publication may outline CA’s general product direction. However, CA may make modifications to any CA product, software program, service, method or procedure described in this publication at any time without notice, and the development, release and timing of any features or functionality described in this publication remain at CA’s sole discretion. CA will support only the referenced products in accordance with (i) the documentation and specifications provided with the referenced product, and (ii) CA’s then-current maintenance and support policy for the referenced product. Notwithstanding anything in this publication to the contrary, this publication shall not: (i) constitute product documentation or specifications under any existing or future written license agreement or services agreement relating to any CA software product, or be subject to any warranty set forth in any such written agreement; (ii) serve to affect the rights and/or obligations of CA or its licensees under any existing or future written license agreement or services agreement relating to any CA software product; or (iii) serve to amend any product documentation or specifications for any CA software product.

Any reference in this publication to third-party products and websites is provided for convenience only and shall not serve as the authors’ or CA’s endorsement of such products or websites. Your use of such products, websites, any information regarding such products or any materials provided with such products or on such websites shall be at your own risk.

To the extent permitted by applicable law, the content of this publication is provided “AS IS” without warranty of any kind, including, without limitation, any implied warranties of merchantability, fitness for a particular purpose, or non-infringement. In no event will the authors or CA be liable for any loss or damage, direct or indirect, arising from or related to the use of this publication, including, without limitation, lost profits, lost investment, business interruption, goodwill or lost data, even if ex-pressly advised in advance of the possibility of such damages. Neither the content of this publication nor any software product or service referenced herein serves as a substitute for your compliance with any laws (including but not limited to any act, statute, regulation, rule, directive, standard, policy, administrative order, executive order, and so on (collectively, “Laws”) referenced herein or otherwise or any contract obligations with any third parties. You should consult with competent legal counsel regarding any such Laws or contract obligations.


Related documents

This section describes how to plan the parameters of the external alarms and complete the data configuration according to the external alarm requirements, through an example...

We have developed the first preliminarily validated screening test in Swahili for speech changes in neurological disorders that is not simply an unadapted translation from English

This study reports on the short term effect of a group drumming intervention programme for aggressive behaviour among adolescent girls (13 to 18 years) diagnosed with conduct

Using the fuzzy theory, the fuzzy synthetic evaluation model are constructed in this study, and then an assessment on the air quality of the selected cities in China are

We propose several collective qualitative decision rules and show that in the context of a possibilistic representation of uncertainty, the use of an egalitarian pessimistic

featured a media campaign, direct access testing through the student health center, sexual health education through residence halls and university courses, and partner referrals

This document utilizes the components within the Danielson Framework for Teaching to provide possible guiding questions for conversations that occur between a principal and