Pelco VideoXpert Enterprise v 3.10 System Design Guide

VideoXpertEnterprise™ v 3.10

System DesignGuide

Document number:C5673M-F

Publication date:09/20

VideoXpertEnterprise™ v 3.10 System DesignGuide

Table of Contents

Understanding the VideoXpert Enterprise™ System 4

Scoping System Licenses 5

Planning for Device Discovery 6

Planning for FISMA/NISTCompliance 7

Selecting Servers 8

Choosing to Use One or More VideoXpert Core(s)™ 8

Choosing to Use One or More VideoXpert Media Gateway(s)™ 8

Planning to Use VideoXpert Aggregation™ 9

Growing theVideoXpertSystem™ 9

Using Active-Active Failover (Single CMG vs Multi-CMG Environments) 10

Clustering Core Servers 10

Working with Clusters 11

Aggregating Systems 11

Using LDAP Authentication 11

Planning for Multi-System Access 12

Understanding Operator Workspace Topology 12

Selecting Recorders 14

Using VideoXpert Storage™ (VxStorage™) for Recording 14

Using VSM, NSM5200, and NSM5300 Servers as VideoXpert Recorders™ 14

Planning for Recording 15

Correlating Recording Storage Platforms and Their Requirements 15

Understanding Camera Support 16

Planning for Camera SDCard Data Retrieval 16

Planning for Auto-Backfill 16

Planning for Recording Schedule Capabilities 17

Using Volumes and Volume Groups 17

Using External NAS Storage (Archive Volume Groups) 17

Using VSM and NSM5300 Models as VideoXpert™ Recorders 18

Planning for Redundant Recording 18

Configuring VideoXpert Storage™ Failover Recording 18

Planning a BackupStrategy for Recorder Database Entries 20

Understanding VideoXpert Storage Failover and Redundant Recording 20

Planning for Database Backups, Redundant Recording, and Recovery 20

Selecting Independent Backup Storage 21

Evaluating Additional System Components 22

Selecting a Load Balancer 22

C5673M-F | 09/20 2

VideoXpertEnterprise™ v 3.10 System DesignGuide

Using VideoXpert Internal Load Balancing 22
Using an External Load Balancer 22

Using a Network Time Protocol (NTP) Server 22

Selecting a DHCP Server 23

Supporting DNS 23

Designing Export Archive Storage 23

Understanding Network Operation Modes 24

Using Rendezvous Points (RP) 24

Using PIM Modes for Multicast Routing 24

Using PIM Dense Mode (PIM-DM) 24
Using PIM Sparse Mode (PIM-SM) 25
Using Sparse-Dense Mode (PIM-SDM) 26

Using DVMRP for Multicast Routing 26

Planning Network Traffic Flow 27

Addressing Traffic and System Limitations 27

Addressing Client-Side Display Limitations 28
Understanding Secondary and Tertiary Stream and Camera Settings 28

Determining Streaming Delivery 28

Making VxOpsCenter™ Streams Adjustments 29
Planning for MJPEG Video Streaming 29
Planning for 4KSupport for Videos, Cameras, and Monitors 30
Planning for H.264 and H.265 Streaming 30

Planning to Use Additional Features 31

Planning for SNMP Monitoring 31

Planning to Use Maps™ 31

Planning to Use Integrations and Plugins 31

Planning for Event Reporting, Logs, and Notifications 33

Understanding Event Types 33

Using Reports 33

Locating Logs 34

Planning for Notifications 34

Appendix A:Video Streaming Diagrams 35

Multicast Recording, Multicast Viewingof a PelcoCamera 35

Unicast Recording, Multicast Viewing (With a VXSProxy)of a PelcoCamera 36

Unicast Recording, Multicast Viewingof a PelcoCamera 37

Unicast Recording, Multicast Viewing of an ONVIFCamera 38

Unicast Recording, Unicast Viewingof a PelcoCamera 39

Appendix B:NetworkProtocols and Ports Reference 40

C5673M-F | 09/20 3

VideoXpertEnterprise™ v 3.10 System DesignGuide

Understanding the VideoXpert Enterprise™ System

VideoXpert is a video management solution designed to fit surveillance operations of any size. Whether
your operation has 100 cameras or 10,000, VideoXpert presents a solution to display, record, and manage
your video resources. But VideoXpert Enterprise is more than a VMS. It is an enterprise-level video and
data management solution designed to combine input from multiple systems, for a cohesive, real-time
understanding of events taking place in your environment.

Action:Prior to obtaining a quote for or purchasing a system, contact a Pelco Technical Sales
Engineer to validate your system configuration.

C5673M-F | 09/20 4

VideoXpertEnterprise™ v 3.10 System DesignGuide

Scoping System Licenses

VideoXpert is licensed for the system, for upgrades, and by channel—the video streams you view and
record. It comes with one (1) license to start. The demo license provides unlimited channels that are active

a period of 60 days. These are active only the first time you install the software, or if the software was

for
pre-installed, the first time you start up the system. In order for the system to function beyond the
evaluation period, add the appropriate quantity of licenses to the system.

Action:Ensure that you have enough licenses. See the current version of the VideoXpert Enterprise
Product Specification for available SKUs. Contact a Pelco Sales Representative for more
information.

C5673M-F | 09/20 5

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for Device Discovery

When adding devices to VideoXpert, the system issues a discovery message and then listens for devices
for up to five minutes. Using the discovery process, VideoXpert adds and commissions your VideoXpert
devices and Pelco and many third-party cameras.

The discovery process will allocate the necessary licenses to commission discovered devices. If the
system does not have the necessary licenses to commission the discovered devices, VideoXpert will add
discovered devices to the system but leave them decommissioned. You must then manually commission
devices either by adding more licenses or decommissioning devices of a similar type to reclaim the
necessary licenses.

Action:Ensure that there are enough licenses for all of the devices on the system, and ensure that
there will be enough bandwidth available to perform Device Discovery without interfering with system
operation.

C5673M-F | 09/20 6

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for FISMA/NISTCompliance

VideoXpertEnterprise is compatible with current FISMA/NIST requirements. If your organization must
comply with these requirements, establish a Risk Management Framework which includes:

l Categorizing the system

l Selecting security controls

l Implementing security controls

l Assessing the system

l Authorizing the system

l Performing continuous monitoring

The current version of the VideoXpertEnterprise™ Configuration Guide for FISMA/NIST Environments
includes configuration guidance and information needed to build a system documentation package for
security control assessments. Specifically, the guide will help organizations documenting the system
through the RMF process with information about how to categorize the system, which NIST-based security
controls are applicable, and how VideoXpert Enterprise implements NIST-based security controls. Testing
of these controls has also been performed by Pelco with VideoXpert Enterprise installed on a FISMA
representative system to ensure functionality under a secure configuration with DISA STIG rules applied.

Note:NIST security control baseline has many allowances for organization-defined settings. While
the VideoXpertEnterprise™ Configuration Guide for FISMA/NIST Environments describes an RMF-
friendly implementation for Pelco VideoXpertEnterprise including NIST security controls and DISA
STIG rules, it might not precisely match your organizationally-defined settings.

Action:To configure your system for FISMA/NISTcompliance, contact Pelco Professional Services
to obtain the VideoXpertEnterprise™ Configuration Guide for FISMA/NIST Environments.

C5673M-F | 09/20 7

VideoXpertEnterprise™ v 3.10 System DesignGuide

Selecting Servers

VideoXpert requires both Core™ and Media Gateway™ servers. Although you can leverage separate Core
and Media Gateway servers for systems of sufficient scale, most systems can easily support servers
running both the Core and Media Gateway (CMG) services. A single CMG server provides the complete
range of VideoXpert functionality that you would expect for systems with fewer than 2000 cameras and 100
simultaneous users.

However, for environments that are especially large, require exceptional redundancy, or incorporate a high
number of low bandwidth and aggregated users, you might install individual Core and Media Gateway
servers.

Action:Determine whether to use a CMG or separate VideoXpert Core server(s) and Media
Gateway server(s) based on the number of cameras and users on your system. See the following
sections for details.

Choosing to Use One or More VideoXpert Core(s)™

VideoXpert Core™ is the heart of the VideoXpert System™, it maintains the database of cameras,
recording devices, users, and permissions. Core works with VxToolbox, from which you can configure and
manage the system. Through VxToolbox, you can administer user accounts and permissions, determining
the system functions and devices users can access. You can create and assign “tags” to quickly organize
cameras and devices within the system. You can also configure and respond to events within the system.

Choosing to Use One or More VideoXpert Media Gateway(s)™

The VideoXpert Media Gateway™ routes video traffic to appropriate users as requested. The Media
Gateway:

l Routes the video to the workstation in a multicast environment

l Accesses the video in a unicast environment

l Transcodes the video for low-bandwidth connected VxOpsCenter clients

You can set the communication method, unicast or multicast, from the Video Source to the Media
Gateway, and from the Media Gateway to the client. The media gateway is capable of transcasting
multicast from the source to unicast for the client, and from unicast to multicast.

Like Core servers, Media Gateways can be added to VideoXpert modularly. You can add Media Gateways
to the system as the media delivery needs increase.

C5673M-F | 09/20 8

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning to Use VideoXpert Aggregation™

VideoXpert Enterprise™ with Aggregation™ allows for expansion at any level of your security environment.
Your system begins with a single server running Core and Media Gateway software. Your system can use
dedicated VxOpsCenter Clients to view live and recorded video, or it can use VxPortal™, which fully
utilizes HTML5 browser technology to deliver a similarly rich experience with no client software required.
As your surveillance needs grow, you can add servers to expand modularly within a single environment, or
you can aggregate multiple VideoXpert Enterprise systems to provide a single point of access for
distributed video management networks.

Caution:Although VideoXpert Professional™ and VideoXpert Enterprise systems can be
aggregated into the same Enterprise system, it is not recommended that you have cross-aggregation
between multiple systems simultaneously.

Growing theVideoXpertSystem™

If your system grows to support additional users, cameras, or sites, or you just want to provide redundancy
within your VideoXpert system, you can separate your VideoXpert Core and Media Gateway servers, and
increase system capacity by clustering servers. If you are using aggregation, you can also aggregate other
VideoXpert systems.

Consider using separate Core and Media Gateway servers when:

l There is a high number of simultaneous users.

l The system must scale to a high number of cameras and users.

l You have high expectations for availability and redundancy.

C5673M-F | 09/20 9

VideoXpertEnterprise™ v 3.10 System DesignGuide

The table below lists typical deployment scenarios, with the maximum number of cameras and concurrent
system users for each deployment; these numbers represent the limits at which the system becomes
unusable (high latency in control requests).

Deployment Cameras Users Availability Additional Requirements

Single CMG 2500 100 Not fault tolerant N/A

Dual CMG 2500 100 Active-Active single

N/A

failover

Triple CMG 10000 500 Active-Active single

N/A

failover

Single Core/ Gateway 3000 200 Not Fault Tolerant N/A

Multi-Core / Gateway >10000 >500 High Availability Independent load balancer

The table presents absolute maximums for VideoXpert deployment scenarios. Your experience might differ
based on your network configuration, network equipment, average video bitrates, and other criteria.

Action:Build the system with at least 10% additional capacity (in terms of cameras, users, or
preferably both), to ensure that the system is responsive and has additional capacity to take on new
users or cameras. When planning a VideoXpert deployment, contact Pelco to ensure that the system
has the capacity to support your environment and needs.

Using Active-Active Failover (Single CMG vs Multi-CMG Environments)

While a single CMG can host nearly 2500 cameras and 100 concurrent users, the system is not at all fault
tolerant; anything that could bring down the server will interrupt access to VideoXpert.

Action:Pelco recommends that if video is mission critical to the business, you build a system with
at least two CMG servers.

Clustering Core Servers

In your VideoXpert environment, Core or CMG servers host the database. Clustering your Core or CMG
servers provides redundancy and enables VideoXpert to scale.

In a clustered environment, each Core or CMG server in the cluster hosts a complete copy of the
VideoXpert database.

In addition, you can install a copy of just the database on a server.

Note: VideoXpert itself performs all of the cluster configuration automatically during the setup
process. If configuring a system containing more than three servers, contact a Pelco Sales
Representatives or a Pelco Technical Sales Engineer.

Within each Core/CMG cluster, one server acts as the primary and the other servers operate as
secondaries. The primary server processes all write operations and pushes data to the secondaries.
Secondary servers replicate the primary server’s database asynchronously.

l If you have two servers and one is unavailable, there is no loss of functionality.

l If you have three servers and one is unavailable, there is no loss of functionality.

l If you have three servers and two are unavailable, the available server is put into a read-only state.

C5673M-F | 09/20 10

VideoXpertEnterprise™ v 3.10 System DesignGuide

In a read-only state, users can still call up video, but would be unable to apply bookmarks, export
investigations, apply tags, and perform other similar operations within the system.

Working with Clusters

A clustered environment requires at least two VideoXpert Core, Media Gateway, or CMGservers.

l Cores and Media Gateways must be on the same VLAN. They must also have static IPaddresses,

and these IP addresses must be different from each other.

l Traffic will be managed by a single Core; if that Core fails, another Core will perform the

management tasks. Other tasks, such as export processing, are shared among all Cores.

l A single Media Gateway will receive streaming requests, but will redirect streaming to other Media

Gateways to balance the load.

l The Media Gateway trans-casts to suit the network topology and needs. While the system is

configured to get multicast streams from sources and to issue multicast streams to clients, you can
select the appropriate communication method both from sources to the Media Gateway and from
theMedia Gateway to clients. The network topology and need for users to access sources
simultaneously will inform your choice.

Use VxToolbox to configure clusters. See the current version of the VideoXpert Toolbox™ Operations

Manual section titled Adding Systems.

Aggregating Systems

VideoXpert Enterprise with aggregation includes an aggregation server, through which you can provide
centralized access to a series of VideoXpert member—VideoXpert Professional and/or VideoXpert
Enterprise—systems. Through the VideoXpert Enterprise server acting as the aggregation system, you
can access and control settings and video for distributed VideoXpert systems.

When adding a member to the aggregation server, you will select your connection speed to the aggregation
server. Your connection speed determines both the performance of video within the VideoXpert
environment hosting the aggregation server and the number of video streams you can reasonably expect to
get simultaneously from the aggregated site.

At present, you cannot change settings for aggregated systems from the VideoXpert instance hosting the
aggregation server. You must change settings for member systems from the member itself.

The aggregation server does not inherit permissions, roles, or users from aggregated members. If
aggregating a VideoXpert Enterprise environment containing roles with restricted permissions, you must re­create these roles and permissions with resource restrictions with the VideoXpert System acting as the
aggregation server.

Using LDAP Authentication

You can configure VideoXpert to validate user credentials from an LDAP server. While the system can
validate credentials over LDAP, you must create corresponding user IDs and roles within the VideoXpert
database against which to validate the credentials. These IDs and roles must match the IDs and Groups in
the LDAP directory exactly (including capitalization) in order for the authentication to pass through. Using
the LDAP interface DOES NOT alter the schema of the LDAP directory, so all permissions to the
VideoXpert system(s) must be defined in the VideoXpert system

You can select the authentication method and parameters used.

l VideoXpert Authentication

When using VideoXpert Authentication, you can set passwords to expire at specific intervals, or to
never expire.

C5673M-F | 09/20 11

VideoXpertEnterprise™ v 3.10 System DesignGuide

l LDAP authentication using simple bind authentication

When using LDAP authentication with simple bind, you can set passwords to expire at specific
intervals, or to never expire.

l LDAP authentication using two-stage binding

When using LDAPauthentication with two-stage binding, you can set passwords to expire at
specific intervals, or to never expire.

l (Optional) If you select LDAP authentication, you can also retrieve users and roles from LDAP

l LDAPauthentication using single sign-on (SSO)

– SSO allows users to log in to multiple systems using a single set of login credentials.

– SSOcan be used with either Single-Stage or Two-Stage binding, and can be used with the

Synchronize Users and Roles From LDAP option.

– SSOrequires valid certificates; each user must have a valid certificate that the system can

access.

– When using LDAPauthentication with SSO, you cannot set passwords to expire. Password

expiration is controlled by the LDAP database policies.

Note:If VxOpsCenter is running on the same server as the VxPro or Core system, SSOwill not be
available. This is due to MSWindows limitations.

Planning for Multi-System Access

You can design the VideoXpert system to run using Single Server Access mode or Multi-Server Access
(MSA) mode. MSA mode enables access to multiple stand-alone VideoXpert systems simultaneously.

When the system is in MSA mode, and Multi-System Access opens without initial credentials is selected
in VxToolbox, users can sign-in to VxOpsCenter without signing-in to a specific system. Credientials will
be required when you select a VXSystem.

Action:When defining the system connections in the VxOpsCenter client software, you can fine­tune the maximum network bandwidth to allow from the system. This will ensure that the video
streams to the workstation have the best resolution and image rate possible without over-saturating
the network link(s) between the system and the user workstation.

Action:When planning to use MSAmode, ensure that the same user ID and passwords exist and
exactly match on all of the different systems. When using MSA mode, Pelco recommends that you
use LDAPto synchronize users and roles to each VXsystem. This will ensure that the username and
passwords match.

Understanding Operator Workspace Topology

The VxOpsCenter, running on an 8-monitor rackmount Workstation, supports up to eight monitors: two
connected directly to the Workstation, and the other six driven independently by Enhanced Decoders. The
Enhanced Decoders enable each monitor connected through a workstation to display up to 25 video
streams while maintaining a seamless user experience.

C5673M-F | 09/20 12

VideoXpertEnterprise™ v 3.10 System DesignGuide

Enhanced Decoder-driven monitors operate just like native monitors; users can move windows across
monitors seamlessly. But, when the user requests a video stream or plugin, the Enhanced Decoder
communicates directly with VideoXpert servers to get and decode video. This enables operators to
maximize the display-capabilities of the VxOpsCenter without complicating the user experience.

Because the Enhanced Decoders do not run the VxOpsCenter application themselves, you might
experience better performance in quantity and responsiveness of HD streams on decoder-driven monitors,
instead of directly-connected monitors.

VxOpsCenter also supports Shared Display mode which provides monitor-wall functionality for a VX
Workstation or a Shared Display Decoder. Monitor walls are specific groups of monitors that are frequently
viewed or used together. A monitor wall does not require an 8-monitor workstation, and it can be scaled as
large as needed using configured monitor numbers. (Monitor numbers are configured in VxOpsCenter.)
Users can send tabs and video to the shared display and control the shared display remotely. In order to
connect to the VXSystem from VxOpsCenter, the user must have permissions that allows the user to add
monitors to the system.

If you require additional monitors on a monitor wall, Shared Display Decoders can be used in combination
with Enhanced Decoders to create a complete monitor wall experience.

C5673M-F | 09/20 13

VideoXpertEnterprise™ v 3.10 System DesignGuide

Selecting Recorders

VideoXpertEnterprise supports multiple recording platforms. In general, traffic and video delivery operates
much the same using any of the available recording platforms. If building a new system, you would
typically use VxStorage E-Series™, VxStorage T-Series™, and VXS5300™ recorders; if migrating from an
existing Endura system, you can use NSM5200 and NSM5300 servers as recorders.

Action:Determine the recorders you will need for your system.

Using VideoXpert Storage™ (VxStorage™) for Recording

VxStorage is a RAID 6, dual power supply, high-availability recording software platform that captures
recorded video for your VideoXpert system. The VxStorage E-Series, VxStorage T-Series, and VXS5300
have the following features:

l Through the user of camera drivers, the VxStorage can support most camera models from most

camera vendors.

l VxStorage supports hot-standby failover configurations so that it can be configured for high-

availability.

l VxStorage has a flexible scheduling engine which allows you to extend the retention of data on the

system without sacrificing video quality.

l On the VxStorageE-Series and VxStorage T-Series, the operating system is contained on an dual

SSD RAID 1 array; hard drives belonging to the RAID array are hot-swappable.

l The VxStorage E-Series, VxStorage T-Series, and VXS5300 servers come with redundant power

supplies.

l Server management uses the embedded iDRAC port which provides out-of-band diagnostics and

remote access to the server OS in the event of a failure.

l VxStorage natively supports IP cameras via ONVIF S, G, Q, and T, and via native driver packages.

You can configure storage through VxToolbox, making it easy to set recording schedules and assign
cameras to your storage servers. Storage supports motion, alarm, and bump-on-alarm style recording, so
you can always capture events relevant to your environment at high quality. You can set different retention
times per camera or camera group. Storage also supports redundant recording by assigning cameras to
multiple recorders, ensuring that VideoXpert continues recording video even if a single storage server falls
offline.

Using VSM, NSM5200, and NSM5300 Servers as VideoXpert Recorders™

VideoXpert supports VSM, NSM5200, and NSM5300 servers as VideoXpert Recorders™. NSM5200 and
NSM5300 servers provide migration paths to VideoXpert; both can operate as recorders within VideoXpert
while continuing to support their respective Endura environments.

The VSM, NSM5200, and NSM5300 recorders can be configured as a pool, where one unit acts as the pool
manager and the other units will take the camera assignments from the pool manager. This also allows for
automatic camera redistribution across the remaining pool members, in the event that one of the pool
members fails. The manager role is handled as an election process within the pool and will be reassigned to
another pool member if the pool manager fails.

C5673M-F | 09/20 14

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for Recording

Recorder types and the settings for recordings can vary widely as VideoXpert systems expand and
change. Devices assigned to recorders will record based on the schedule and recording triggers you
configure through VxToolbox.

Factors to consider when planning for recording include:

l The number and type of data sources (video, audio, PTZ vs. fixed, etc.) that will be added to a

recorder

l The recording behaviors (triggers)

l The data source recording schedules

l The global maximum retention period of recordings (The retention period for individual devices must

be shorter than the global retention period.)

l The transmission method (multicast or unicast)

l The stream(s) to record (primary, secondary, and/or tertiary)

l Whether the system uses motion recording

l Whether the system uses bump-on-alarm recording

l Whether and how many data sources are using auto-backfill of recording gaps

l The maximum bitrate of recordings

l How many bookmarks are expected to be stored, the standard retention limit of unlocked

bookmarks, and the expected retention of locked bookmarks

Action:Ensure that you have enough recorders (standard and failover), recording space (per recorder
and system-wide), and appropriate network capacity to support the likely storage requirements and
network traffic.

Correlating Recording Storage Platforms and Their Requirements

VideoXpert supports the recording platforms identified in Table 1: Recording platforms.

Table 1: Recording platforms

NSM5300

(2.4.3 or

later)

VxPortal /
Pelco
Utilities

250 Mbps
in; 32
streams
out (per
pool)

Parameter

Maximum

VxStorage

T-Series

144TB 288 TB 48 TB 48 TB 48 TB

VxStorage

E-Series

VXS5300 VSM

capacity

RAIDlevel RAID6 RAID6 RAID 6 RAID6 RAID 6

Configure using VxPortal VxPortal VxPortal VxPortal / Pelco

Utilities

Bandwidth

1 Gb: 700 Mbps in;
175 Mbps out

10 Gb: 1000
Mbps; 175 Mbps
out

1 Gb: 700 Mbps in;
175 Mbps out

10 Gb: 2500
Mbps; 175 Mbps
out

450 Mbps in; 175
Mbps out

250 Mbps in; 32
streams out (per
pool)

C5673M-F | 09/20 15

VideoXpertEnterprise™ v 3.10 System DesignGuide

Note:Optera cameras or cameras supporting high framerates and resolutions require an increased

amount of storage. This is mostly due to increased bitrates; in some cases, increased processing
load might also become a factor.

Understanding Camera Support

Different VideoXpert systems support different cameras, as shown in Cameras supported per VideoXpert

System.

Table 2: Cameras supported per VideoXpert System

Camera VideoXpert Storage VSM/NSM

Optera Yes Yes

1

EVO, GFC-Multi, Esprit Ti, Sarix Value Yes No

Other Pelco Yes Yes

ONVIF-compliant Profile S No

Native APIsupport Axis, Hikvision, Panasonic, Vivotek, etc. No

1

Optera is supported on the VSM/NSM however the camera is limited to Tile mode. The Panomersive

modes are not available.

Planning for Camera SDCard Data Retrieval

Some cameras have SD card data retrieval capabilities to populate the recorded video. Using this
capability can cause extra network traffic due to the SD card polling and the recorder replenishment from
the SD card to the recorder(s).

Action:Consider this capability when devising the recording schedule(s), and accommodate for the
potential extra network bandwidth needed.

Planning for Auto-Backfill

For cameras that have local recordings, you can enable Auto-backfill recording gaps using on-camera

storage. When selected, if there is a gap in recording of a camera (an edge device), the recorder will query

the camera for recordings, and automatically download video and audio (if present) to fill the gap.

You can also adjust the following settings:

l The number of cameras from which data is downloaded at one time

Downloading from more cameras uses more bandwidth, which will slow the data transmission rate.

l A time-interval to apply if edge devices (cameras) and the recorder lose communication

The VXSystem will attempt to connect to the edge device again at the time-interval specified in this
field.

l The number of attempts to reconnect if edge devices and the recorder lost communication

l If available, the maximum bitrate at which to download data

Action:Determine the number of cameras that will use Auto-Backfill, the time-intervals to apply
when reconnection is necessary, the number of attempts to reconnect, and the maximum bitrate for
each download. Use this information to ensure you have enough network bandwidth to accommodate
the reconnection attempts and downloads.

C5673M-F | 09/20 16

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for Recording Schedule Capabilities

The primary recording schemes used by VideoXpert are:

l Continuous recording can be performed at full or reduced frame rates.

l Event-triggered recording (including alarm, analytics, and motion detection events) is performed at

full frame rate.

l Bump-on-alarm is a type of continuous recording during which the system records reduced frame

rate (I-Frame only) video during normal situations, and records full frame rate video during an alarm
or event.

l In VxToolbox, administrators reduce the frame rate of previously recorded video after a specified

number of days, resulting in increased retention time and storage cost savings. To use this feature,
see instructions in the current version of the VxToolbox OperationsManual.

Action:Full frame-rate recordings will consume more storage than reduced frame rate recordings.
Ensure that your system has enough storage to accommodate the expected recording types.

Using Volumes and Volume Groups

You can organize your device video storage by creating and managing Volumes and Volume Groups.

l A volume is a logical directory in which you want to store video.

l A volume group is a group of volumes to which cameras are assigned and distributed. You can use

volume groups to:

– Separate types of storage (like internal vs. external).

– Set different retention parameters for different sets of drives.

– Write video to more than one volume. When all volumes are full, the system will overwrite

volume containing the oldest stored video.

l The system ships with a volume group called Default Volume Group. You can rename or delete this

volume group.

l An archive volume group is a volume group to which the recorder will move the oldest video from the

other volume groups, instead of deleting the oldest video. See Using External NAS Storage (Archive

Volume Groups) for more information about the archive volume group.

Action:To use volumes and volume groups, determine which cameras will be assigned to each, on
which recorder the volumes and volume groups will reside, and where the archive volume group will
reside. Ensure there is adequate storage and network bandwidth to support their use.

Using External NAS Storage (Archive Volume Groups)

By connecting an external volume (network storage/NAS) to your system, you can extend your retention
time for VideoXpert Storage recorders. When your VideoXpert Storage recorder achieves its maximum
capacity and would normally begin to delete the oldest video, it will send video over to the NAS instead.
Video will still adhere to retention parameters, even when moved over to external storage. The experience
in accessing video is the same, whether a recording is served from a the VxStorage recorder or an external
server.

To use external storage, the external storage server reside on the VideoXpert network, and should not have
login credentials. If the server requires login credentials, VxStorage will not be able to transfer video to the
external server.

C5673M-F | 09/20 17

VideoXpertEnterprise™ v 3.10 System DesignGuide

Action:As video transfers from a VxStorage recorder to an external storage server, bandwidth of

your incoming cameras is equal to the bandwidth out to external storage. When using external
storage, plan storage distribution to ensure bandwidth availability for incoming cameras, storage
overflow, and user impact in viewing recorded video.

Action: While each VxStorage recorder can only have a single archive group, multiple VxStorage
recorders can use the same NAS server. In this case each VxStorage recorder must point to a
different path/folder on the NAS server; pointing multiple VxStorage recorders to the same archive
group network path will cause video to expire earlier than expected and without warning.

Using VSM and NSM5300 Models as VideoXpert™ Recorders

Through VxToolbox, you can set camera associations and recording schedules; however, you must
configure other aspects of your recording devices through other interfaces. Configure storage pools through
the VSM/NSM5300 Web interface. Configure recording schedules through Pelco Utilities.

Planning for Redundant Recording

Redundant recording can be configured to perform in the following ways:

l Two recorders recording the primary stream

l Two recorders recording the same camera with one recording the primary stream and the other

recording an alternate, lower-quality stream

Recordings do not have to be the same size/retention unless those are part of the end-user's standards.

Plan for the following network considerations when configuring redundant recording:

l Multicast recording of the streams will allow for only a single stream coming out of the camera, no

matter how many recorders are recording the stream.

l Unicast recording of the cameras will have implications in that most cameras can only push out 2-4

copies of the stream; if two streams are in use by recorders, then there might not be another stream
available for live viewing.

l Double the number of network ports you would otherwise need for recorder connection.

l In Unicast-recorded networks, the bandwidth handled by the uplinks between the switches and the

switch backplanes will be doubled; this could result in resource over-allocation or recording/viewing
failures.

l Unicast recording could result in the need to create LACP teamed network trunk uplinks, or

necessitate replacement of the switches/switchports with 10Gig inter-switch network ports.

When planning HVAC and UPS for rendundant recording:

l Double the number of power plugs you would otherwise need.

l Plan for additional BTU mitigation.

l Double the number of network ports you would otherwise need for recorders.

l Include enough UPS battery capacity to maintain the desired power-outage run time for the system

including the additional recorder(s) power load.

Configuring VideoXpert Storage™ Failover Recording

By putting VxStorage units in failover mode, you can assign VxStorage models to act as hot-standbys for
live recorders and ensure uninterrupted recording when active recorders fail. Pelco recommends that you

C5673M-F | 09/20 18

VideoXpertEnterprise™ v 3.10 System DesignGuide

use this feature to implement a failover strategy that ensures full time recording and video availability in
your environment in the event of possible network, power, or hardware failures.

A single designated failover recorder can monitor up to eight active recorders. When a recorder is in failover
mode, if there is a failure in any of the recorders that the failover unit is assigned to monitor, the failover unit
will take over and continue recording the camera groups and schedules previously belonging to the failed
recorder.

Failover recording does not ensure access to historical data from failed recorders; it only ensures continued
recording. If you want to maintain uninterrupted access to recorded video, even when a recorder fails, then
record cameras redundantly across multiple recorders.

A designated failover recorder operates one-to-one with the recorders it monitors; if a unit fails, it will take
over for the cameras and schedules belonging to the failed unit. If the failover unit is already engaged and
another recorder fails, those cameras will go unrecorded until you correct the issue.

The failover recorder does not need to be of the same capacity as the production units, but that they must
be sized to accommodate the video retention that would be needed until the production recorder can be
brought back online.

Action: Ensure that you have enough standby VxStorage units to support your failover requirements.

C5673M-F | 09/20 19

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning a BackupStrategy for Recorder Database Entries

A system backup contains the system database, including the previous 30 days’ events. Backups do not
capture exported video or any settings that you changed outside of VxToolbox (for example: changes made
directly to configuration scripts). The speed of each backup depends on the size of the VideoXpert
database, network bandwidth, and other variables.

VideoXpert Storage takes database recovery points daily, and stores these points for eight days. You can
also initiate a manual backup. Pelco recommends that you back up your system after initial setup, when
you have configured a significant number of users and devices, and after significant changes to your
system configuration.

Action: Ensure that an appropriate backup strategy is in place, and that there is enough storage and
network bandwidth to support the strategy.

Understanding VideoXpert Storage Failover and Redundant Recording

Recorders are not part of the standard Core/CMG cluster. However, typical deployments should arrange
for some measure of redundancy to ensure that required streams and video never go unrecorded.
VideoXpert Storage (VxStorage)supports two methods for redundancy:

l Redundant recording—assigning the same camera to multiple recorders, or setting the camera

primary stream to record on one recorder and the secondary or tertiary stream on another recorder

l Failover monitoring—assigning a recorder to act as a hot-standby for one or more recorders

Redundant recording is a quick way to ensure video uptime and redundancy, but requires an equal amount
of storage across each recorder set to record a stream.

Putting recorders in Failover mode allows you to assign them to monitor and act as a hot-standbys for
active recorders; if any of the active recorders fail, the failovers will record the camera groups and
schedules in place of the failed recorders ensuring nearly uninterrupted recording.

Note: During a failover, you might experience a recording gap of up to 30 seconds.

Planning for Database Backups, Redundant Recording, and Recovery

Because VideoXpert Storage takes database recovery points daily and stores recovery points for eight
days, record cameras redundantly if you might need to recover more than eight days of video in the event of
a failure.

You can also have the database as a service on a server that is separate from the VideoXpert Core and
Media Gateway servers. This enables you to store additional copies of the database on other servers.

Action:If you plan to record redundantly, ensure that you have enough storage on the system and
enough network bandwidth to manage the traffic associated with making all of the required backups.

C5673M-F | 09/20 20

VideoXpertEnterprise™ v 3.10 System DesignGuide

Selecting Independent Backup Storage

For any system with more than one VideoXpert Core, you must backup to a network location. The network
location must be accessible via a UNC path (for example:\\backupserver\backups). Local and mapped
network drives (S:\backups) are not supported for the backup process. Pelco recommends that you store
backups to a server independent of other VideoXpert hardware, preferably containing its own RAID array,
ensuring that your backups are safe from catastrophic failures.

Action:If your VideoXpert system has more than one Core, select one or more independent
servers for backup storage. Plan to accommodate network traffic for these servers.

C5673M-F | 09/20 21

VideoXpertEnterprise™ v 3.10 System DesignGuide

Evaluating Additional System Components

Depending on the shape and scale of your system, you might need one or more of the following
components.

Selecting a Load Balancer

For load balancing, you can use VideoXpert Internal Load Balancing or an external load balancer.

Using VideoXpert Internal Load Balancing

If VideoXpert internal load balancing is used, then the CMGs must be on the same VLAN. This is because
the high availability VIP is owned by the CMGs .

Using an External Load Balancer

For systems with more than three CMGs or multiple independent Core and Media Gateway servers, you
might need an external load balancer. The load balancer provides the virtual IP address used to target
clustered servers.

Your load balancer must meet the following requirements.

l High-level requirements:

– HTTP

– HTTPS

– websocket support

– RTSP support; alternatively, support for raw TCP/IP connections

– Support for multiple sets of backend servers; necessary if using separate Core and Media

Gateway servers

– Support for application-level HTTP healthchecks

– Support for application-level RTSP or TCP/IP health-checks

l High availability requirement—Must be able to configure load balancer appliances such that if an

appliance becomes unavailable (for example:through power loss), the other appliance(s) can
assume functionality of the unavailable appliance.

l Recommended Features:

– Allow administrators to gracefully disable servers during upgrades

– SSL offloading

Action:Select a load balancer that is appropriate for the size of your system. Contact Pelco
Professional Services for assistance.

Using a Network Time Protocol (NTP) Server

All servers in your VideoXpert system must reference a time server to ensure that all devices belonging to
the system use the same time. Time disparities may result in errors when recording and recalling video.

l Pelco recommends that you use an official NTP server or purchase and use a network clock to keep

your system synchronized.

C5673M-F | 09/20 22

VideoXpertEnterprise™ v 3.10 System DesignGuide

l You can choose to use the VideoXpert Core cluster on your VideoXpert system as the NTPserver,

but doing so will allow the time of your system to drift significantly from the actual time if there is no
time source.

Action:Pelco strongly recommends that you have a dedicated NTP server for VideoXpert. Other
options include:network routers configured as NTP servers, an application like Nettime for internet
connected systems, and a GPS or cellular network time server device.

Selecting a DHCP Server

Pelco recommends that you use a DHCP server to assign and manage addresses for the devices within
your VideoXpert network, using one of the following:

l Microsoft Windows with DHCPServer

l DHCPservices on a network switch

l A separate, dedicated DHCP server (for advanced installations with multiple ranges and a need to

edit address reservations.)

Action:Select an appropriate DHCPserver to accommodate the number of ranges on the
VideoXpert System and whether you will edit address reservations.

Supporting DNS

On systems that are connected to networks with DNS servers, the devices can resolve the hostnames of
the source devices, rather than just relying on the IP addresses. This is key for multiple NTP server
connectivity. Using DNS also provides the administrator the ability to:

l Leverage the DNS system to create DHCP IP address reservations.

l Resolve the name of the device using the IP addresses.

Designing Export Archive Storage

You can store exports in a network location on or off of the VideoXpert Core server. Storing exports off of
the Core ensures greater availability to exported video, especially in clustered environments. When storing
exports on the Cores in a clustered environment, exports are not shared among Cores; each export is only
stored on the Core server on which it was created. If the server storing an export fails, users lose access to
the exported video; if the server fails and you have to restore from a backup, you will lose your exports on
that Core server. Storing exports in a separate location ensures availability independent of any individual
Core server, and allows you to easily backup video exports at whatever interval is most convenient for you.

You can either save exports to the default location or to another location. Any other location must be
defined in URIformat; it cannot use the local drive letter format. If export encryption is enabled, a password
is required for saving to the default location.

Action:Design an appropriate storage strategy that will ensure the availability of exports. Plan for the
amount of storage and the network traffic necessary to support the strategy.

C5673M-F | 09/20 23

VideoXpertEnterprise™ v 3.10 System DesignGuide

Understanding Network Operation Modes

Using Rendezvous Points (RP)

PIM-SM builds a shared multicast distribution tree within each domain, and the RP router is at the root of
this shared tree. Although you can physically locate the RP anywhere on the network, it must be as close
to the source as possible. Only one active RP router exists for a multicast group.

At the RP router, receivers meet new sources. Sources use the RP to identify themselves to other routers
on the network; receivers use the RP to learn about new sources.

The RP performs the following tasks:

l Registers a source that wants to announce itself and send data to group members

l Joins a receiver that wants to receive data for the group

l Forwards data to group

Sample RProuter configuration is as follows:

ip multicast−routing

ip pim send−rp−announce loopback0 scope 16

ip pim send−rp−discovery scope 16

interface loopback0

ip address <address> <mask>

ip pim sparse−dense−mode

interface ethernet0

ip address <address> <mask>

ip pim sparse−dense−mode

interface serial0

ip address <address> <mask>

ip pim sparse−dense−mode

Using PIM Modes for Multicast Routing

Protocol Independent Multicast (PIM) routing operates in either Sparse Mode (SM), Dense Mode (DM), or
Sparse Dense Mode(SDM).

Action:Before selecting a PIM operating mode, consider the impact that protocol selection will have
on the network.

The following sections provide an overview of PIM modes and use considerations.

Using PIM Dense Mode (PIM-DM)

PIM-DM is easier to install than PIM-SM. The network engineer will enable PIM-DM on each network router
that is required to route multicast traffic. PIM-DM operates in what is referred to as a push model. Traffic is
initially flooded to all neighbors that have formed a PIM neighbor relationship. Downstream routers will then
determine if the traffic is necessary and either forward the traffic appropriately or send a prune message to
an upstream router to suppress the flow of multicast traffic. Keep in mind that although the traffic has been
suppressed, the (S,G) state is still maintained in the multicast routing table. One of the major drawbacks to
PIM-DM is that multicast routing switches that are not actively transmitting a multicast flow might still be

C5673M-F | 09/20 24

VideoXpertEnterprise™ v 3.10 System DesignGuide

required to maintain that state. Maintaining this state can lead to the consumption of additional resources
on the switch even though no active client on that router has requested the multicast traffic. During the
flood and prune cycle (S,G), states are flooded to every multicast router on the network and every multicast
router will maintain the (S,G) state as long as the multicast source is actively transmitting. The resulting
traffic flow for multicast will follow the shortest path tree (SPT) from source to receiver.

l Determine if the Layer 3 routing devices support state refresh. Because PIM-DM will flood

traffic throughout the network to build (S,G) states in each downstream multicast router, careful
consideration must be given to the support of state refresh. Multicast routing devices that support
state refresh will prevent periodic flooding. PIM-DM operates in a flood and prune cycle. The
multicast routing tree is flooded every three minutes and relies on pruning mechanisms to determine
whether or not downstream routers require the multicast traffic. Periodic flooding of the network can
be a major concern for networks for which bandwidth is limited. Layer 3 devices that support state
refresh prevent the countdown timer on the (S,G) entry from expiring. If the countdown timer never
expires, the multicast source will no longer flood the network periodically after the initial flood cycle.

l Determine the multicast table routing table entry limitations of each switch on the network.

There is a finite limit for each switch concerning the number of multicast routing table entries the
switch can handle. If the available multicast routing table entries are exhausted, further entries
might fail to be allocated to the table resulting in a multicast group that can no longer be routed. As a
network engineer, you must ensure that the switch that is being used is not exceeding its capacity
for the multicast routing tables. Pelco has a list of recommended switches that have been tested
with respects to its multicast routing table capacity. It is the responsibility of integrators or network
engineers to contact the switch manufacturer to assess the capabilities of the switch and any
limitations with respect to multicast routing table entries.

l Select recommended network switches or test non-recommended switches. In addition to the

multicast routing table, a selected switch must be able to handle an adequate number of IGMP
entries. Switch manufacturers specify the number of IGMP entries a switch can handle. When
switches exceed these limits, they typically will either flood or block multicast traffic. Pelco
maintains a list of recommended switches that have been tested for their maximum recommended
IGMP entries. If an integrator or network engineer selects a switch that is not from the
recommended switch list, it is the responsibility of the integrator or network engineer to contact the
vendor to determine the IGMP limitations of the switch selected.

l Verify network limitations associated with wireless connections. Due to the limited bandwidth

associated with wireless connections, PIM-DM might not be an appropriate selection. The flood and
prune cycle might result in a wireless network link that becomes saturated.

Using PIM Sparse Mode (PIM-SM)

While PIM Sparse Mode requires careful consideration during the design process, there are major benefits
associated with using PIM-SM as opposed to PIM-DM. Unlike PIM-DM, PIM-SM has a dedicated RP to
send messages to build both the shared (*,G) and source (S,G) sides of the tree. The end result is that PIM­SM will not perform flood and prune cycles to build trees for forwarding multicast traffic. When the multicast
traffic is not flooded to all PIM-enabled devices, devices not in the path of transmission will not maintain
entries in the multicast routing table. This will result in lower utilization of switch resources that are not in
the SPT.

Due to the operation of PIM-SM, placement of the RP can be a critical decision in network design. If a
centralized RP is selected for all traffic in the network, that switch must be able to handle the appropriate
number of multicast routing table entries for all traffic traversing the network. As an alternative, you can use
multiple RPs that serve as candidates for multicast routing. Filtering can be implemented to distribute the
multicast routing load across multiple RPs. This type of application allows you to distribute the multicast
routing load across multiple PIM-SM routers and, if designed properly, isolates multicast traffic to intended

C5673M-F | 09/20 25

VideoXpertEnterprise™ v 3.10 System DesignGuide

segments of a network. For example, if a multicast recording network storage pool is implemented and the
RP also serves as the local designated router, multicast recorded traffic would use its local designated
router as the RP and isolate the majority of the multicast flows to the local router. Since the SPT is local to
the switch, multicast recording traffic would be contained within a segment of the network.

In an implementation using PIM-SM, only the initial video packets are sent to the RP. If a single RP is used
in a network, after the encapsulated video in the register message is sent, all remaining video packets use
the SPT from source to destination.

An SPT threshold can be configured to force a multicast flow to bypass the SPT. Care should be taken if
SPT thresholds are to be modified.

If a single RP is used in PIM-SM, it is critical that the multicast routing switch have enough resources to
handle all (*,G) and (S,G) entries that will be created in the multicast routing table. Even though the traffic is
traversing the SPT, resources must be allocated to handle all existing multicast routing table entries, and
any processing of joins and prunes throughout the network. Packet replication, RPF recalculation, state
maintenance, and register processing all create memory and CPU loads on the RP. Depending on the size
of the network and scalability requirements, different Layer 3 devices might be selected as RP based on
their resources.

The default response of PIM on some switches is to fall back to PIM-DM in the event that a RP cannot be
found. Based upon the network topology this might or might not be a desired effect. Always take into
account the effect that reverting to PIM-DM might have on the network. This response is present on Cisco
systems.

Using Sparse-Dense Mode (PIM-SDM)

Some implementations of PIM simultaneously support Dense Mode for some multipoint groups and Sparse
Mode for others. This provides the regularity of flood/prune broadcasts in Dense mode and the bandwidth
savings of Sparse mode.

Using DVMRP for Multicast Routing

DVMRP is a routing protocol supporting multicast transmission. Stemming from Routing Information
Protocol (RIP) and used in the Internet multicast backbone (Mbone), DVMRP allows for tunneling multicast
messages within unicast packets. It also supports rate limiting and distribution control based on destination
address, and it is responsible for the following tasks:

l Routes multicast datagrams

l Periodically floods multicast traffic (similar to PIM-DM)

l Allows use of non-multicast aware edge devices

Note: When choosing PIM-DM or DVMRP as a multicast routing protocol on systems that include
wireless devices or that require remote access to the system, understand that these protocols have
bandwidth limitations that are negatively affected by periodic flooding of data streams.

C5673M-F | 09/20 26

VideoXpertEnterprise™ v 3.10 System DesignGuide

MEDIA GATEWAY

OPS CENTER CLIENT

ENTERPRISE SYSTEM

VSM/NSM5200

LOAD BALANCER/NLB

CORE

HTTP/HTTPS

RTSP

RTP

NSM

Pelco API

VX Storage

Planning Network Traffic Flow

Command and control traffic (user actions within the system), occur over HTTP or HTTPS depending on
your system. You can configure workstations to operate over HTTP or HTTPS, and you can select the port
for communications with VideoXpert Servers. (For information on ports, see the section titled Appendix

B:NetworkProtocols and Ports Reference.)

Video is delivered to clients either via RTSP or RTP, depending on the Media Gateway Communication
method for which your system is configured. When configured for unicast delivery to clients, the Media
Gateway re-streams video to the client. When configured for multicast delivery to clients, video streams
directly from the camera, encoder, or RTSP source from which the stream is requested.

Addressing Traffic and System Limitations

VideoXpert systems are tested to determine how many users and cameras a system supports before
performance degrades significantly. The systems tested represented environments with strong network
connections using VSM models for storage. For purposes of performance tests, “users” are simultaneous
operators performing continuous, expected duties, including streaming video, receiving events, controlling
(PTZ) cameras, and exporting video.

Refer to the current version of the VideoXpert Enterprise Product Specification for details.

C5673M-F | 09/20 27

VideoXpertEnterprise™ v 3.10 System DesignGuide

For help determining the best system size and configuration to meet your needs, contact a Pelco Sales
Representative.

Addressing Client-Side Display Limitations

As aVxOpsCenter workstation uses an increasing amount of memory, users might experience “jittery”
mouse controls on decoder-driven monitors. This problem is most likely to occur if users are running
applications in tandem with the VxOpsCenter client, or is viewing a large number of Optera or HD streams
on locally-connected monitors.

Understanding Secondary and Tertiary Stream and Camera Settings

VxOpsCenter uses step-down behaviors when under heavy load. Optimal camera configuration ensures
that you always view the highest possible quality video and prevents the system from entering I-Frame­only or disconnection step-down scenarios.

l Set secondary streams to 640 x 352 (or the corresponding 4:3 equivalent, depending on available

aspect ratios) at 5 images per second or lower to ensure secondary stream performance in a 4x4
layout.

l For Optera cameras, set the I-Frame interval to 6 and use Smart Compression on the camera to

reduce the bandwidth being used. Optera cameras consume more system resources than other
cameras.

l Set tertiary streams to the lowest resolution and frame rate that is acceptable to ensure tertiary

stream performance in high-density layouts and for use with low-bandwidth connections.

See the section titled Determining Streaming Delivery.

Determining Streaming Delivery

When the system is in Multicast mode, the stream comes directly from the camera. When the system is in
Unicast mode, stream delivery is largely the responsibility of the Media Gateway, and is dependent on the
communication method you select, your connection to the system, and the connection of aggregated
devices to the system. In general, you can receive streams identified in the table below.

Stream quality Factors

Primary Bandwidth > 100 Mbps; Cell size≥ 25% of layout

Secondary Cell size < 25% of layout; quality step-down when VxOpsCenter is under heavy

load

Tertiary Quality step down (from secondary stream) when VxOpsCenter is under heavy

load

C5673M-F | 09/20 28

VideoXpertEnterprise™ v 3.10 System DesignGuide

Stream quality Factors

I-Frame Only Quality step down (from tertiary stream) when VxOpsCenter is under heavy load

MJPEG/H.264/H.265

VxOpsCenter:

- Client bandwidth > 5 mbs will support the full range of options, lower connections
will be forced to MJPEG.

- Aggregation connections with bandwidth >= 10 mbs will support the full range of
options, lower connections will be forced to MJPEG.

- PTZ Camera, when in MJPEG mode, will only support click-to-center. A warning
is given.

VxPortal: Same as VxOpsCenter, but without any warnings provided regarding
MJPEGlimitation.

See Appendix A:Video Streaming Diagrams for examples of streaming configurations and parameters.

Action:Because the system delivers secondary streams as a step-down adjustment, configure
secondary streams with reduced resolution and frame rate compared to the primary. Similarly,
because the system delivers tertiary streams as a step-down adjustment, configure tertiary streams
with reduced resolution and frame rate compared to the secondary.

Making VxOpsCenter™ Streams Adjustments

By default, VxOpsCenter displays streams based on the size of the cells in a tab. Typically, cells that
represent 25% of a layout or more display full resolution, primary streams; cells smaller than 25% of a
layout deliver secondary streams. The 2x3 layout is also designed to show primary, full resolution streams
in all cells.

When the CPU load on the decoding device (Workstation, Shared Display Decoder, or Enhanced Decoder)
that displays the streams crosses the 75% threshold, VxOpsCenter will request lower-quality streams in
an attempt to reduce CPU load below 50% utilization. Stream quality is reduced from primary to secondary,
and secondary to I-Frame-only mode until the CPU load drops below 50%. If the reduction in stream quality
does not reduce the CPU load below the 50% threshold, VxOpsCenter will begin disconnecting streams,
citing limited CPU resources.

Planning for MJPEG Video Streaming

The system streams MJPEG video as described in Determining Streaming Delivery. MJPEG Video
includes the following limitations:

l PTZ controls for MJPEG streams are limited to click-to-center controls.

l MJPEG streams might not synchronize perfectly when in sync playback mode. This is due to the

nature of the MJPEG pull mechanism. You can still use the sync playback controls to manipulate
video, but synchronization might be noticeably affected.

l When Media Gateway CPU utilization reaches its threshold, it will reject requests to transcode new

MJPEG streams.

Action:If cameras will be expected to return MJPEG video at greater than two I-Frames per
second, ensure that there is enough bandwidth on the system and that the Media Gateway(s) can
accommodate the MJPEG streams.

C5673M-F | 09/20 29

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for 4KSupport for Videos, Cameras, and Monitors

The VideoXpert system supports 4K monitors, 4K live and playback modes for cameras that support
streaming through ONVIF, and the ability to export recorded 4K segments.

Enhanced Decoders do not support 4Kvideo output to monitors. You can connect 4Kmonitors to
Enhanced Decoders, but the monitors will step down to show 1080p only.

Action:To support 4K video, cameras, and monitors, ensure that the VideoXpert system has enough
storage and network bandwidth to support the greater amount of data that will be stored and
streamed.

Planning for H.264 and H.265 Streaming

The system streams H.264 and H.265 video for users with connections greater than 10 Mbps, or cameras
belonging to aggregated sites with connection speeds greater than than 5 Mbps.

Note:VxPortal cannot consume H.265 streams directly, so uses MJPEG.

Cameras configured to use H.264 and H.265 streaming will consume more network and WANbandwidth
that MJPEG and other streaming methods.

Action:Determine how many devices will be using H.264 or H.265 streaming and ensure adequate
network and WANbandwidth.

C5673M-F | 09/20 30

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning to Use Additional Features

Planning for SNMP Monitoring

If you use anSNMPmanager to monitor the VideoXpert environment (native hardware or your own
hardware), you must monitor the same events as the VideoXpert Front PanelService. VideoXpert Front
Panel Service is a software utility that runs on Pelco hardware (native) that monitors the VideoXpert
services.

Pelco has other management information bases (MIBs) that can be used in SNMPmonitoring on the
VideoXpert system.

At a minimum, monitor and send traps for the following software services:

l VideoXpert Core

l VideoXpert Exports

l VideoXpert Media Gateway

l VideoXpert OpsCenter Communications

l VideoXpert Storage

l VideoXpertStorageDatabase

Pelco recommends that you monitor and send traps for the following hardware events:

l Running software (executable name, path, and status)

l Processor load

l Storage/memory statistics

l Hard disk failures

l Power failures

l System uptime

Action:Ensure that you have enough network bandwidth to support communication of SNMPtraps
you plan to send.

Planning to Use Maps™

Maps™ imports and uses AutoCAD 2013 DWG files, raster maps (jpeg or png), and ESRI Street) maps,
allowing customers to use their pre-existing building maps with Maps in VideoXpert.

Action:If the system will use ESRIStreet files, ensure that each station that uses them has internet
access.

Planning to Use Integrations and Plugins

VideoXpert supports integrations both through VxOpsCenter and directly to VideoXpert Core. Integrations
through VxOpsCenter function as plugins—applications that a user can add to VxOpsCenter—either
operating in conjunction with or overlaying video. Integrations through Core typically inject events into
VideoXpert, extending the functionality of the system and issuing events to VxOpsCenter users when the
third-party system records them. Through integration, you can add incident reporting, analytics, license
plate recognition, and other features to VideoXpert. You can download VideoXpert integrations and plugins
from https://www.pelco.com/partners/technical-partners/.

C5673M-F | 09/20 31

VideoXpertEnterprise™ v 3.10 System DesignGuide

Pelco offers integration tools and software development kits (SDKs) to help you extend the functionality of
your VideoXpert system through the Pelco Developer Network at https://www.pelco.com/training-

support/developer-support/. Pelco’s Partner First site provides information about and access to a number

of ready-made integrations and plug-ins for VideoXpert.

Action: Find out storage and network requirements from the manufacturer of each plugin that you will
use. Ensure that the VideoXpert System has enough storage space and network bandwidth to support
all aspects of each plugin.

Action:Obtain the appropriate license(s)from Pelco or the third-party for each plugin that you use.

C5673M-F | 09/20 32

VideoXpertEnterprise™ v 3.10 System DesignGuide

Planning for Event Reporting, Logs, and Notifications

You can pull reports (in CSV format) containing events and configuration information to audit your system
and VideoXpert users. Reports requiring a start date and time include an event history; reports that do not
require a start date and time provide current configuration information only. System Administrators can
schedule automatic report generation in VxToolbox.

In addition, each VideoXpert product produces and stores its own logs, which roll over every 30 days. You
can use the logs to troubleshoot issues, or to help Pelco troubleshoot issues you might encounter in the
field.

In addition to the on-screen notifications, the VideoXpert system can be configured to send notifications
using SMSmessaging or via email. (SMS and email are not available on isolated network VideoXpert
systems.)This will increase network traffic based on the number and frequency of notifications that are
expected.

The following sections provide more information about events, reports, logs, and notifications.

Action:Ensure that you have enough storage and network bandwidth to support storing and
retrieving/accessing the reports and logs, and enough bandwidth to manage the network traffic
associated with notifications. For SMSnotifications, ensure network connections outside the system
through the Firewall. For email notification, ensure access to an internet-connected mail server.

Understanding Event Types

Event types typically refer to who or what is the cause of an event, and do not necessarily determine the
report in which an event will appear:

l Admin events typically occur at the administrative level.

l Analytic events are the result of software analytics.

l Client events are reported by the client.

l External events are injected into the system by a third party.

l Hardware events are the result of physical hardware issues and sensor readings.

l System events are the result of normal operator actions.

l Custom event types are configured through the Rules Engine in VxToolbox.

By default, events expire every 30 days. However, you can choose to keep events for up to 90 days, and
you can set different expiration periods for both events local to the system and events from aggregated
sites.

You can also set a maximum number of events; the default maximum is 10,000 events.

Using Reports

You can create and generate reports, and export them from the system. Reports are exported in CSV
format.

Report Templates preloaded to VxToolbox include:

l Default Camera Report

l Default Device Report

l Default Event Report

l Default Recording Gap Report

l Default Role Report

C5673M-F | 09/20 33

VideoXpertEnterprise™ v 3.10 System DesignGuide

l Default Camera Role Access Report

l Default Storage Report

l Default User Report

l Default User Action Report

l Default System Status Report

You can edit an existing report template, or create a new one and edit the template details to include only
the information needed. This will affect the storage space necessary to accommodate the reports.

Locating Logs

Each VideoXpert product produces and stores logs to assist in troubleshooting. Logs are available in the
following locations:

l Core—C:\ProgramData\Pelco\Core\logs

l Media Gateway—C:\ProgramData\Pelco\Gateway\logs

l VxStorage—C:\ProgramData\Pelco\Storage\logs

l VxOpsCenter—C:\ProgramData\Pelco\OpsCenter\Logs; logs are available within the application

by navigating to About VxOpsCenter and clicking Get Logs.

Each of these location requires storage for the anticipated number and size of the logs.

Planning for Notifications

System notifications are delivered via a pop-up window on your VxOpsCenter Client. The System
Administrator can configure settings including:whether there is a pop-up banner associated with the
notification, whether the notifications require acknowledgment, and whether it will play a sound. The
configuration must define the individual groups that will be notified on each specific event type.

Notifications can also be sent using SMSmessaging or via email.

l SMS messaging requires bandwidth and network traffic connectivity outside the system through

firewalls. To send SMS notifications, you must create a Twilio account for the system and then the
system Administrator must configure the notification in VxToolbox.

l E-mail notification can only be used if there is a connection to an e-mail server via SMTP which

requires a connection to an internet-connected mail server. E-mail notifications are not available for
isolated network VX systems.

Configure event details and notifications to ensure that the right users are notified when the system records
a particular action or alarm. The way in which rules are configured determine how many notifications are
likely to be sent and how often.

C5673M-F | 09/20 34

VideoXpertEnterprise™ v 3.10 System DesignGuide

Appendix A:Video Streaming Diagrams

Multicast Recording, Multicast Viewingof a PelcoCamera

Live—Live video is authenticated by the CMG and streamed from the camera/network to the viewing client
and the VXS.

Playback—Playback is always unicast from theVXS.

*Bandwidth calculated is for one camera being viewed in a single screen on one monitor.VideoXpert
intelligently selects camera streams based on multi-screen configuration and client utilization.OPS
Viewing clients subscribe to camera source IGMP groups to view video in multicast.

C5673M-F | 09/20 35

VideoXpertEnterprise™ v 3.10 System DesignGuide

Unicast Recording, Multicast Viewing (With a VXSProxy)of a PelcoCamera

Live—Live video is authenticated by the CMG and streamed from the VXS to the viewing client. This is the
default on VX v 2.5, but is configurable onVXv 3.0 and later.

Playback—Playback is always unicast from theVXS.

*Bandwidth calculated is for one camera being viewed in a single screen on one monitor.VideoXpert
intelligently selects camera streams based on multi-screen configuration and client utilization.OPS
Viewing clients subscribe to camera source IGMP groups to view video in multicast.

C5673M-F | 09/20 36

VideoXpertEnterprise™ v 3.10 System DesignGuide

Unicast Recording, Multicast Viewingof a PelcoCamera

Live—Live video is authenticated by the CMG and streamed from the camera/network to the viewing client.
This is the default on VXv 3.0 and later, and can be changed so that theVXS provides the multicast
stream.

Playback—Playback is always unicast from theVXS.

*Bandwidth calculated is for one camera being viewed in a single screen on one monitor.VideoXpert
intelligently selects camera streams based on multi-screen configuration and client utilization.OPS
Viewing clients subscribe to camera source IGMP groups to view video in multicast.

C5673M-F | 09/20 37

VideoXpertEnterprise™ v 3.10 System DesignGuide

Unicast Recording, Multicast Viewing of an ONVIFCamera

Live—Some ONVIF cameras are re-streamed by the VXS to the OPSViewing client and authenticated
byVideoXpert Core; some are re-streamed by the Media Gateway to the viewing client. Features are
camera-dependent.

Playback—Playback is always unicast from theVXS.

*Bandwidth calculated is for one camera being viewed in a single screen on one monitor.VideoXpert
intelligently selects camera streams based on multi-screen configuration and client utilization.OPS
Viewing clients subscribe to camera source IGMP groups to view video in multicast.

C5673M-F | 09/20 38

VideoXpertEnterprise™ v 3.10 System DesignGuide

Unicast Recording, Unicast Viewingof a PelcoCamera

Live—Live video is authenticated by the CMG and streamed from the Media Gateway to the viewing client.
This is the default on VXv 3.0 and later, and can be changed so that theVXS provides the multicast
stream.

Playback—Playback is always unicast from theVXS.

*Bandwidth calculated is for one camera being viewed in a single screen on one monitor.VideoXpert
intelligently selects camera streams based on multi-screen configuration and client utilization.OPS
Viewing clients subscribe to camera source IGMP groups to view video in multicast.

C5673M-F | 09/20 39

VideoXpertEnterprise™ v 3.10 System DesignGuide

Appendix B:NetworkProtocols and Ports Reference

The tables below correspond to VideoXpert software components; in some cases, the same server can
host multiple components. The ports listed must be open to ensure VideoXpert functions properly.

Table 3: VideoXpert Core™ protocols and ports

Protocol Ports Service

TCP/HTTP 80 HTTP, used for camera configuration as necessary

TCP/HTTPS 443 HTTPS

UDP/SSDP 1900 SSDP discovery target on 239.255.255.250

TCP/Hazelcast 6001 Hazelcast communications, before cluster

configuration

TCP/Hazelcast 6002 VxDatabase Hazelcast communications, before

cluster configuration

TCP/Postgres 15432 Database

TCP/Hazelcast SSL 16011 Hazelcast SSL communications, after cluster

configuration

TCP/Hazelcast 16012 VxDatabase Hazelcast TLS communications, after

cluster configuration

Table 4: VideoXpert Media Gateway™ protocols and ports

Protocol Ports Service

UDP/RTP/RTSP All When streaming Unicast from cameras, the Media

Gateway uses ports in the range 41950-65535 to
receive the data, but the client may request data on
any port. Multicast data will use the port configured
on the camera, thus all UDP ports must be available.

TCP/RSTP 554 RTSP

TCP/HTTPS 5443 Internal API, HTTPS (MJPEG and other

communication)

TCP/Hazelcast 6002 Hazelcast communications, before cluster

configuration

TCP/HTTP 8090 Internal API, HTTPS (MJPEG) not used by the

system

TCP/Hazelcast SSL 16002 Hazelcast SSL communications, after cluster

configuration

C5673M-F | 09/20 40

VideoXpertEnterprise™ v 3.10 System DesignGuide

Table 5: VxStorage™ protocols and ports

Protocol Ports Service

UDP/SSDP 1900 SSDP discovery

TCP/RTSP 5544 RTSP video and audio command and control

TCP/Hazelcast 6003 Hazelcast communications

TCP port/ HTTP

9091 HTTP, APIcalls

TCP/HTTPS 9443 HTTPS, APIcalls

UDP/RTP 41950-65535 Receiving streamed video and audio

UDP/RTCP 41950-65535 Receiving media streams metadata

If recording in multicast mode, the required UDPport range is determined by the cameras.

Table 6: VSM / NSM5200 / NSM5300 Storage™ protocols and ports

Protocol Ports Service

TCP/SSH 22 SSH, remote shell connectivity

UDP/DHCP 68 DHCP Client

TCP/HTTP 80 HTTP, web access

UDP/SNMP 161 SNMP

UDP/SNMPTRAP 162 SNMP-trap

TCP/SMUX 199 SNMPmultiplexing service

TCP 1605 NSTERM Server

UDP 1781 NSTERM Search, recording services

UDP/UPnP 2900-2901 UPNP Discovery

TCP/HAL 4343 HAL

TCP 10000-10008 Private services for NSM5300/VSM

TCP 25556 DB Replication service for redundant recorders

UDP/UPnP 32768-61000 UPNP Communication

TCP/UPnP 49152 UPNP Communication

C5673M-F | 09/20 41

VideoXpertEnterprise™ v 3.10 System DesignGuide

Table 7: Enhanced Decoder™ protocols and ports

Protocol Ports Service

TCP/FTP 21 Used for firmware updates, snapshot downloads,

etc.

UDP 4500-4900 Unicast streaming via Windows host (not decoders)

TCP 43241 Proprietary control protocol

Enhanced decoders operate on Xubuntu 14.04 and communicate over the protocols identified in the table
above.

Table 8: Shared Display™ Decoder / VideoXpert Workstation™ protocol and ports

Protocol Ports Service

TCP/VNC 5900-5906 Allows the decoders to operate in concert with the

workstation as a single desktop/workspace

This is the normal port range for a workstation with 5 enhanced decoders. The actual range may vary. The
range can be confirmed on the workstation using the installed TightVNC Service Control Interface. If no
enhanced decoders are connected, these ports are not necessary.

Table 9: VxOpsCenter Client™ protocols and ports

Protocol Ports Service

TCP/FTP 21 Provides firmware updates to enhanced decoder,

downloads snapshots and logs from the enhanced
decoder

TCP/HTTPS 443 RESTAPI

TCP/RTSP 554 Video and audio command and control

UDP 4502-4900

1

Unicast streaming

TCP/RTSP 5544 Video and audio command and control

1

Unicast streaming previously used a port range of 4500-4600.

Table 10: Outgoing and general protocols and ports (for VxToolbox™, VideoXpert Core™, Media
Gateway™, VxStorage™, VideoXpert Professional™, and for viewing clients)

Protocol Ports Service

TCP/FTP 21

TCP/HTTP 80

TCP/NTP 123

Time synchronization

TCP/HTTPS 443

C5673M-F | 09/20 42

VideoXpertEnterprise™ v 3.10 System DesignGuide

Protocol Ports Service

TCP/RTSP 554

UDP/SSDP 1900

SSDPdiscovery

UDP/WSD 3702 WS discovery

TCP/UDP/RDP 3389 Remote desktop

UDP 4500-4900 Unicast streaming via Enhanced Decoders

UDP 4502-4900

1

Unicast streaming via Windows host (not Enhanced
Decoders)

TCP/HTTPS 9334 Discovery of VxStorage

UDP/UPnP 49152 UPnPcommunication

TCP/UDP—RTP/

N/A Streaming and control

RTCP/RTSP

1

The VxOpsCenter and Shared Display unicast streaming previously used a port range of 4500-4600.

C5673M-F | 09/20 43

VideoXpertEnterprise™ v 3.10 System DesignGuide

Pelco, Inc.

625 W. Alluvial Ave., Fresno, California 93711 United States

[phone] Tel
[phone] Fax
[phone] International T el
[phone] International F ax

www.pelco.com

Pelco, the Pelco logo, and other trademarks associated with Pelco products referred to in this publication are trademarks of Pelco, Inc.
or its affiliates. ONVIF and the ONVIF logo are trademarks of ONVIF Inc. All other product names and services are the property of their
respective companies. Product specifications and availability are subject to change without notice.

© Copyright 2020, Pelco, Inc. All rights reserved.