EMC Surveillance for VIEVU Body Worn Cameras … · ... Isilon), object store on-premises (Isilon Swift), private cloud storage ... OpenStack Swift APIs, EMC CAS, and ... Isilon scale-out

EMC Solution for VIEVU Body Worn Cameras

Functional Validation Guide H14533

01

Copyright © 2015 EMC Corporation. All rights reserved. Published in USA.

Published October, 2015

EMC believes the information in this publication is accurate as of its publication date. The information is subject to changewithout notice.

The information in this publication is provided as is. EMC Corporation makes no representations or warranties of any kind withrespect to the information in this publication, and specifically disclaims implied warranties of merchantability or fitness for aparticular purpose. Use, copying, and distribution of any EMC software described in this publication requires an applicablesoftware license.

EMC², EMC, and the EMC logo are registered trademarks or trademarks of EMC Corporation in the United States and othercountries. All other trademarks used herein are the property of their respective owners.

For the most up-to-date regulatory document for your product line, go to EMC Online Support (https://support.emc.com).

EMC CorporationHopkinton, Massachusetts 01748-91031-508-435-1000 In North America 1-866-464-7381www.EMC.com

2 EMC Surveillance for VIEVU Body Worn Cameras Functional Validation Guide

Introduction 5

Solution overview ...........................................................................................6Scope..............................................................................................................6

Solution components 7

VIEVU cameras................................................................................................ 8VERIPATROL.....................................................................................................8EMC Elastic Cloud Storage (ECS)......................................................................8EMC CloudArray...............................................................................................9VNX and VSS storage platforms....................................................................... 9Isilon clustered storage system..................................................................... 10

Solution configuration 11

Body-worn camera storage architecture.........................................................12VNX/VNX-VSS architecture............................................................... 12

Body camera connectivity options................................................................. 13Isilon architecture............................................................................ 13CloudArray architecture....................................................................14

Testing and validation 17

Test summary................................................................................................18

Chapter 1

Chapter 2

Chapter 3

Chapter 4

CONTENTS

EMC Surveillance for VIEVU Body Worn Cameras Functional Validation Guide 3

CONTENTS


CHAPTER 1

Introduction

This functional validation guide provides compatibility guidelines for EMC storage arrays and storage clusters.

l Solution overview ...................................................................................................6l Scope......................................................................................................................6

Introduction 5

Solution overviewThe use of body-worn cameras in law enforcement is expanding globally. As lawenforcement agencies grapple with the additional workload and realize the impactrelated to their IT infrastructures, it is critical that they procure validated solutions thataddress the underlying system requirements. EMC's approach to addressing bodycamera system requirements provides an open and flexible architecture for severalstorage strategies relevant to the body worn camera dataflow and workload.

The EMC Surveillance Solutions for Body-Worn devices meet the system requirementsrelated to traditional on-premises storage (VNX, VNX-VSS, Isilon), object store on-premises (Isilon Swift), private cloud storage (ECS, CloudArray), and hybrid cloud storage(CloudArray EMC Cloud SP). Customers can deploy each option independently or incombination to meet a broad scope of storage requirements and ultimately deliver aninfrastructure strategy that meets their immediate needs and can scale to accommodatefuture requirements.

ScopeThis guide provides results from a functional test that was conducted to ensure thecompatibility of VIEVU cameras with EMC storage. The test does not establish sizingguidelines.

This guide is intended for use by internal EMC sales and pre-sales personnel, andqualified EMC and VIEVU partners.

This guide provides functional information regarding VIEVU using EMC storage solutions,such as:

l EMC Isilon ®

l EMC VNX ® and VNX-VSS ®

l EMC CloudArray ® to EMC Elastic Cloud Storage ™ (ECS)

l EMC CloudArray to EMC Isilon Swift

Introduction


CHAPTER 2

Solution components

This chapter provides information about the components configured in this solution.

l VIEVU cameras........................................................................................................ 8l VERIPATROL.............................................................................................................8l EMC Elastic Cloud Storage (ECS)..............................................................................8l EMC CloudArray.......................................................................................................9l VNX and VSS storage platforms............................................................................... 9l Isilon clustered storage system............................................................................. 10

Solution components 7

VIEVU camerasVIEVU cameras are built tough for law enforcement, tactical and military use. VIEVU bodyworn video cameras are easy to wear and use. Use VIEVU to record hands-free video &audio from your point of view.

The following list describes the primary VIEVU cameras:

LE3

l A leading police body worn video camera deployed by over 4000 agencies in 16countries.

l LE3 features include HD video resolution, 16GB internal memory and up to 12hours of record time.

l Video from the LE3 camera is securely stored in the cloud or on-site usingVIEVU's proprietary software.

l A FIPS 140-2 compliant digital signature process is used to verify the video hasnot been altered.

VIEVU²

l A feature-rich, small form factor wearable video camera that is hands free,rugged, and waterproof.

l It streams and stores real time video to your iPhone or Android smartphonewithout the need for a cable.

l Once videos are on your phone, you can edit and upload them to Facebook,Twitter, and YouTube or email, text and store them in your DropBox folder.

l VIEVU² eliminates the need to transfer videos later.

The VIEVU cameras support the following resolutions:

l 1280 x 720

l 848 x 480

VERIPATROLVERIPATROL is an on-site video management system for VIEVU LE cameras. VERIPATROLincludes several layers of security to keep your data secure and utilizes a FIPS 140-2compliant Digital Signature process to prove that video has not been altered. VERIPATROLis fully network capable and can support deployments of all sizes. The video from theVIEVU camera is categorized and stored to an EMC storage solution using VERIPATROL.

EMC Elastic Cloud Storage (ECS)ECS is a complete software-defined cloud storage platform that supports the storage,manipulation, and analysis of video surveillance and unstructured data on a massivescale on commodity hardware. ECS is specifically designed to support the mobile, cloud,and Big Data workloads that are similar to large scale body camera workloads.

ECS provides GUI, RESTful API, and CLI interfaces for provisioning, managing, andmonitoring storage resources. Storage services provided by the unstructured storageengine (USE) ensure that video is available and protected against data corruption,hardware failures, and data center disasters. It enables global namespace management

Solution components


across geographically dispersed data centers and geo-replication. The USE enables thefollowing storage services:

Object serviceLets you store, access, and manipulate video and unstructured data. The objectservice is compatible with existing Amazon S3, OpenStack Swift APIs, EMC CAS, andEMC Atmos APIs.

Hadoop Distributed File System (HDFS)Helps you use your ECS storage infrastructure as a Big Data repository that you canrun Hadoop analytic applications against (in-place).

The provisioning service manages the provisioning of video surveillance storageresources and user access. Specifically, it handles user management, authorization, andauthentication for all provisioning requests, resource management, and multi-tenancy.

You can scale up, scale out, and add users, applications, and services, as well asmanage your local and distributed storage resources for your surveillance data through asingle view.

EMC CloudArrayEMC CloudArray provides cloud-integrated storage that extends high-performancestorage arrays with cost-effective cloud capacity. By providing access to a private orpublic cloud storage tier through standard interfaces (iSCSI, NAS) CloudArray technologysimplifies storage management for video files and provides offsite protection.

CloudArray takes advantage of local storage, also known as cache, which serves twopurposes. The first is to provide local performance for body camera file ingest minimizing,and often eliminating the effects of cloud latency. The second is to serve as a buffer forread/write operations.

CloudArray is offered as both a physical appliance and a virtual appliance that can bedeployed either on site or in a cloud compute environment. CloudArray uses cachestorage for performance acceleration. User defined policies dictate how much data iskept in the local cache. Each CloudArray can manage multiple policies and cachessimultaneously, each of which can be individually configured to support the needs of thedifferent application workloads.

The physical appliance is a dedicated server that comes with built in CloudArray software.The size of the hardware appliance determines the total amount of local cache available.The appliance's storage can be carved into multiple, smaller caches to allow for greatergranularity and flexibility. In a CloudArray physical appliance, the total amount of cacheavailable is pre-determined based on the hardware configuration.

In a virtual appliance however, the cache can be any local storage accessible by theCloudArray software. In fact, a CloudArray virtual appliance can be mapped to multipledifferent storage types based on requirements. For example, a CloudArray virtualappliance can be mapped to solid state drives (SSDs) for a cache associated with high-performance or production systems, and SATA drives for archive or other data with lowerperformance and availability requirements.

VNX and VSS storage platformsThe VNX-VSS (Video Surveillance Storage) series is a purpose-built (block) storagesystem solution for the Video Surveillance workload that brings storage reliability,

Solution components

EMC CloudArray 9

simplicity, and affordability to the distributed environments on your video surveillancenetwork.

The VNX and VNX-VSS series can be used to address both distributed and large-scalecentralized surveillance requirements where block-only (iSCSI, Fibre Channel)connectivity are required. For applications that support block-only connectivity, the VNXline of storage arrays can scale to multiple petabytes for large scale centralizeddeployments.

The VNX series is ideal for surveillance data storage requiring enhanced capabilities andis designed for a wide range of environments that include midtier through enterprise. TheVNX series includes offerings with file-only, block-only, and unified (block and file)implementations.

Both the VNX and VNX-VSS are managed through a simple and intuitive user interfacecalled Unisphere. This single pane of glass integrates information from varied sourcesacross multiple applications and environments into a single display to complete theunified experience.

Isilon clustered storage systemIsilon scale-out network-attached storage (NAS) was designed and developed specificallyto address the needs of storing, managing, and accessing digital content andunstructured data such as surveillance video. An Isilon cluster provides simple dynamicscaling to increase or decrease the total capacity of the Isilon cluster.

An Isilon clustered storage system is composed of multiple nodes. These nodes areintegrated with the Isilon OneFS® operating system, which is a distributed networked filesystem that unifies a cluster of nodes into a single shared resource.

The single namespace across the multinode cluster allows equal access to video files nomatter which node is currently connected. Isilon SmartConnect provides load balancingacross the cluster's nodes to ensure the even distribution of video files.

Solution components


CHAPTER 3

Solution configuration

This chapter provides information about the components configured in this solution.

l Body-worn camera storage architecture.................................................................12l Body camera connectivity options......................................................................... 13

Solution configuration 11

Body-worn camera storage architectureSurveillance data can consume vast amounts of storage while still requiring access toboth live and archived video. The demand for surveillance storage is increasing due tohigher camera resolutions, increased numbers of deployed cameras, and extendedretention times for the video. These factors are forcing the need for greater storageperformance and scale.

VNX and VNX-VSS provide high performance and fault tolerant storage options for videostorage data. Isilon provides file or local object store in highly expandable storageoptions that meet the current and future needs for surveillance data. Both Isilon andVNX/VNX-VSS options provide partners and customers the benefit of locally securedvideo storage.

ECS enabled with CloudArray provides a cost-effective means to meet the current andfuture needs for surveillance data. By deploying CloudArray, partners and customers canreceive the benefits of private or hybrid cloud storage without needing to integratedirectly with the cloud storage platform.

In public safety organizations, or in the public safety quadrants within corporateenvironments, there is an unexpected exponential requirement for storage dedicated tobody-worn camera video. This expansion can be due to a number of factors including:

l New deployment of body worn cameras

l Expansion of existing worn camera deployment

l Increasing body worn camera resolution and/or frame rate

l Increased retention period to meet compliance related to legal proceedings andgovernmental retention policies

The following figure shows the possible body camera connectivity options.

Figure 1 Overview of Body Worn Camera connectivity options

VNX/VNX-VSS architectureEMC Surveillance Validation Labs continually test video management systems (VMS)including multiple body camera vendors ensuring compatibility with EMC local storagesystems and EMC CloudArray with a private cloud and public cloud on multiple cloudservices.

The data flow from the camera to either a VNX or VNX-VSS starts when the body cameratransfers its video to a workstation running a video server.

EMC's Surveillance Validation Labs also functionally validated that body camera videofiles can be stored to an iSCSI-attached VNX or VNX-VSS storage array. After the video is



copied to the array a video server can be used to review video and manage archivedvideo based on preset retention policies.

Body camera connectivity optionsBody-worn devices can be configured using different EMS storage systems, such as VNX/VNX-VSS, Isilon, and CloudArray.

VNX and VNX-VSS (iSCSI)A video server is attached to EMC VNX storage using an iSCSI initiator. The videoserver ingests video files from the body camera client which is stored on the EMCVNX/VNX-VSS storage.

Isilon share (CIFS)The video server mounted shares are created on the EMC Isilon scale-out storagecluster and the video server ingests video files from the body camera client which isstored on the Isilon cluster. Depending on the capabilities of the body cameraapplication, the Isilon shares can be mounted via CIFS (SMB) or NFS.The following figure shows a body camera configuration where a video server storesthe video to the VNX or VNX-VSS array via an iSCSI connection.

Figure 2 VNX/VNX-VSS body camera architecture

Isilon architectureThe data flow from the camera to an Isilon server starts when the VIEVU camera transfersits video to the workstation running VERIPATROL.

The EMC Surveillance Lab also functionally validated that the VIEVU video files can bestored to a SMB attached Isilon scale-out storage device. After the video is copied to thecluster, VERIPATROL can be used to review and manage archived video based on presetretention policies.

An Isilon cluster provides simple dynamic scaling to increase or decrease the totalcapacity of the Isilon cluster, which can determine the amount of storage available toVERIPATROL.

Isilon's OneFS operating system provides a single name space across the multinodecluster. This single namespace design allows equal access to files no matter which nodeVERIPATROL is connected to. Isilon SmartConnect provides the ability to load balanceacross the nodes of the cluster.

The following image shows a body camera configuration in which VERIPATROL stores thevideo to the EMC Isilon through a CIFS or iSCSI connection.


Body camera connectivity options 13

Figure 3 Isilon body camera architecture

CloudArray architectureThe EMC Surveillance Lab tested both ECS Cloud Storage and Isilon Swift object storageenabled by EMC CloudArray. This allows the video server to use both Private Cloud (ECS)as well as basic object storage via Isilon Swift.

Hybrid Cloud functionality was also tested using CloudArray with an EMC Cloud ServiceProvider over a public network. EMC CloudArray enables the use of multiple Cloud ServiceProviders for basic object storage that are compatible with CloudArray without the needfor direct integration.

Platforms tested:

l Elastic Cloud Storage (private cloud)

l Isilon Swift (private object storage)

l AT&T Cloud Object Store (hybrid cloud)

The video server can be connected to EMC CloudArray using iSCSI or CIFS protocols. Bodycamera video is ingested and is stored to CloudArray. CloudArray in turn seamlesslyhandles the write, read, and delete operations to the object store related to normal enduser and application activity.

The data flow from the camera to either a Private Cloud or a Public Cloud starts with thebody camera transferring its video to a workstation or server typically through a client orapplication agent.

The video server mounted shares are created on the EMC Isilon scale-out storage clusterand the video server ingests video files from the body camera client which is stored onthe Isilon cluster. Depending on the capabilities of the body camera application the Isilonshares can be mounted via CIFS (SMB) or NFS.

The following image shows a body camera configuration where VERIPATROL stores thevideo to the Isilon storage cluster through an OpenStack Swift connection.



Figure 4 Local object store, Private Cloud, and Public Cloud vis CloudArray

EMC CloudArray uses a local cache called the CloudArray Cache. The CloudArray Cache iseither direct attached storage (DAS), an iSCSI-attached EMC VNX/VNX-VSS, or an Isiloncluster. When a video server writes to the CloudArray it first writes to cache marking thefile as a dirty page. A dirty page is an indicator that the video file only resides on thecache and that it has not been copied to cloud storage, but will be copied to cloudstorage based on a least recently used algorithm (LRU). This LRU approach ensures thatany video that must be reviewed immediately is available locally avoiding file transfersfrom the cloud. This minimizes network utilization and latency of viewing the videorelated to cloud retrieval; ultimately it can improve the end user experience.

The video is then copied to cloud storage via the clouds API protocol, such as Amazon'sS3 protocol. Our test were conducted using Amazon's S3 protocol. This is true if writing toeither the private cloud or most public clouds.

The video is then copied to cloud storage via Amazon's S3 protocol to EMC ECS or publicclouds that support the S3 protocol such as AT&T. The EMC Surveillance Lab alsofunctionally validated that the video files can be stored to a network attached Isiloncluster when using the Swift protocol. This requires a licensed version of Isilon Swift forthe target cluster. After the video is copied to the cloud's object store, that video existssimultaneously on the cache and the object store, and the dirty page indicator isremoved. CloudArray will eventually purge the cached version of the video. Althoughcloud array works on the block level, conceptually the video file purge occurs when theCloudArray Cache is nearly full and video selected for removal is the oldest video basedits last access.

The size of the cache is based on implementation requirements and cache retentionconsiderations including the rate of file ingest, frequency of video recall, private cloud, orpublic cloud connectivity. We recommend a cache of 12 TB or greater so that it issufficient to accommodate the timeframe that encompasses 95% of the expected videoreviews.

When video is recalled for review, that video is read directly from the cache. If the cachedcopy was removed, the video is recalled from cloud storage and again placed on thecache. Once the video is recalled from the object store to the cache, that video thenresides on the cache as if it was newly accessed.


CloudArray architecture 15



CHAPTER 4

Testing and validation

This chapter describes the testing used to validate this solution.

l Test summary........................................................................................................18

Testing and validation 17

Test summaryThe functional test determined that VERIPATROL was used to successfully review thevideo from the VIEVU camera using EMC storage solutions. For more information aboutserver sizing guidelines, refer to VIEVU system specifications.

The following EMC storage solutions were tested:

l EMC Isilon

l EMC VNX and EMC VNX-VSS

l EMC CloudArray to EMC Elastic Cloud Storage (ECS)

l EMC CloudArray to EMC Isilon Swift

The VIEVU cameras create a file for every 0.99GB of video data. There are two supportedresolutions for video file creation. The details of the two resolutions are as follows:

1280 x 720 resolution

FPS: 30Camera usable space: 14.7GB usable11 files in one charge with 10 file of 0.99GB and the 11th file of 884MBDuration of video data in each file: 24min, 26secCamera storage usage: 3.84GB free, 10.8GB usedFile format: .MP4Video Encoding format: H264

848 x 480 resolution

FPS: 30Camera usable space: 14.7GB usableSix files in one charge with five files of 0.99GB, and the 6th file of 580MBDuration of video data in each file: 52min, 19secCamera storage usage: 9.14GVB free, 5.56GB usedFile format: .MP4Video Encoding format: H264

With CloudArray, we tested playback with the video coming from the CloudArray cacheand then from the object store. When reading video from the Object Store the video isbrought first to the Cache and then the CloudArray serves that video to the requester.There is no meaningful difference in recall time between reading from cache and readingfrom the Object Store on a local private cloud.

Our testing discovered no appreciable differences in latency when retrieving video fromthe CloudArray Cache as compared to retrieving video from ECS or Isilon Swift on aprivate cloud on our lab network.

Testing for the public cloud scenario has provided longer video recall times from theprivate cloud, but recall times may vary within a public cloud. The design of public cloudarchitecture includes too many factors to derive a video recall time that can beuniversally applied.

The VERIPATROL retention time feature was successfully tested on all storage options andtesting scenarios including local storage, private cloud storage, local object storage, andhybrid cloud storage.



NOTICE

The information in this guide should be used only as a compatibility guide and not as aperformance baseline for sizing purposes.


Test summary 19



Documents

EMC Surveillance for VIEVU Body Worn Cameras … · ... Isilon), object store on-premises (Isilon Swift), private cloud storage ... OpenStack Swift APIs, EMC CAS, and ... Isilon scale-out