Deploying sipXecs is easy and installation is usually done in hours. Because of its Web based administration interface as well as the ability of sipXecs to put the users in control, ongoing system administration is easily done by a receptionist or similar person. The following examples are intended to give some typical and therefore recommended system configurations. They range from very small deployments at small offices with a few users up to very large and therefore fully redundant deployments serving several thousand users.

 
sipXecs is designed to be the ideal enterprise IP voice solution for any office, starting at 4 to 12 users with 4 analog trunk lines up to several hundred users in a single location using one or several digital trunk lines (T1 / PRI). It's hardware cost are low and operating expenses are minimized by leveraging low-cost computer servers, commodity hardware, and Linux.  sipXecs connects to the telephone network (PSTN) using traditional analog (FXO) or digital (T1 / E1 / PRI) trunks. Alternatively sipXecs can leverage SIP trunking services provided by a service provider (ITSP). This configuration requires a SIP-aware firewall or Session Border Controller (SBC) to securely traverse the Enterprise firewall and NAT. Possible solutions we tested include SBCs from Ingate or Intertex. Connecting remote workers either requires a corporate Virtual Private Network (VPN) or an SBC capable of offering services to remote workers. A native sipXecs based solution for SIP trunking and remote workers is in progress (project sipXbridge).

sipXecs can scale up easily using more powerful servers or by distributing its components over more than one server. Such configurations can serve up to several thousand users per location and are also suited for multi-branch deployments. Branch offices can have different configurations: a) The branch office runs its own instance of sipXecs, b) The branch office uses the corporate IP network to connect to sipXecs running in a central location, and c) Each branch office can have a redundant or non-redundant configuration of sipXecs. Each branch office can have local gateways. This can improve resiliency for emergency calls, offer least-cost routing, or off-load the corporate WAN network connecting calls directly to the telephone network locally.

sipXecs provides a complete solution to your enterprise telephony needs. The system’s architecture lets you easily distribute servers, gateways and intelligence strategically on your network, within one office or among branch offices, for cost savings, high reliability, backup and load balancing. Capabilities such as automatic trunk fail-over and redundancy, high-availability, and least cost routing are must-have features in today’s environment.

Bundles are given below for the following scenarios:

• Small office with 4 to 12 users using analog trunk lines
• Small office with 12 to 25 users using analog trunk lines
• Mid-size office up to about 100 users using a single T1 / E1 digital trunk
• Mid-size office up to about 200 users using a dual T1 / E1 digital trunk
• Mid-size office up to about 400 users using a quad T1 / E1 digital trunk
• Large office up to about 1000 users using 8 T1 / E1 digital trunks
• Large office up to about 5,000 users 

Recommended phones for both desk phones, attendant consoles and softphones are valid across all deployments . Capabilities of phones vary and the recommended list given below is based on high quality phones with excellent support in sipXecs (meaning they are tested across different firmware releases). Phone support is based on simple plugins. If the phone you like is either not on the list or not well supported it is fairly easy to fix. 

The reason for recommending Audiocodes gateways is simple: sipXecs supports plug & play management for Audiocodes gateways, which greatly simplifies deployment. In addition we have done the most testing with Audiocodes gateways and they are therefore known to work. Audiocodes does not pay SIPfoundry to advertise its gateways. Other gateways known to work are from Cisco or Patton Electronics, although for now they need to be manually configured. Other SIP compliant gateways most likely will work too, but have not been as extensively tested.

Deployment for a small office with between 4 and 12 users: 

Hardware: Small Form Factor PC (e.g. Dell Optiplex or similar) with 1 GB RAM. Standard server without any special HW or add-on cards. Audiocodes 4 port FXO (MP114) Audiocodes 4 port FXS (MP114) Description: Supports small office with IP telephony incl. unified messaging, voicemail, auto-attendant, group paging and intercom, remote worker solution, desktop integration, FAX and conference phone support. Installs in hours. Grows with your business by adding gateways up to several hundred users.

Deployment for a small office with between 12 and 25 users: 

Hardware: Small Form Factor PC (e.g. Dell Optiplex or similar) with 2 GB RAM. Standard server without any special HW or add-on cards. Audiocodes 8 port FXO (MP118) Audiocodes 4 port FXS (MP114) Description: Supports small office with IP telephony incl. unified messaging, voicemail, auto-attendant, group paging and intercom, remote worker solution, desktop integration, FAX and conference phone support. Installs in hours. Grows with your business by adding gateways up to several hundred users. There is no limit on the total number of active calls other than LAN bandwidth. Trunk fail-over redundancy is available for all systems as a standard features. A minimum of two trunk gateways are required. System redundancy is optional (see below).

Deployment for a mid-size office with between 100 and 400 users: 

A typical user to trunk ratio is somewhere between 3 and 4, meaning that for up to 100 users a single T1 (23 trunk lines) or E1 (30 trunk lines) should be sufficient unless you run call center or conferencing applications.  Up to about 400 users are typically served with 4 T1 or E1 digital trunks. A single Audiocodes Mediant 1000 gateway supports up to 4 T1 or E1 trunk interfaces and additional analog FXS (station side for FAX or analog phones) or FXO analog trunk ports.

Hardware: 1U or 2U Server (e.g. Dell, Intel or similar) with 2 to 4 GB RAM. Standard server without any special HW or add-on cards. Audiocodes Mediant 1000 modular gateway (alternatively a Mediant 2000 or TP-260 PCI card can be used) Description: Supports mid-size offices with IP telephony incl. unified messaging, voicemail, auto-attendant, remote worker solution, desktop integration, FAX and conference phone support. Installs in hours. Grows with your business by adding gateways.  Scales to multi-branch office deployments. Other gateways, analog or digital, can be added as needed where they are needed. This solution does not include redundancy, which is available as an option and often recommended for systems larger than 50 users. Trunk fail-over redundancy is available with a minimum of 2 trunk gateways.

Deployment for a mid-size to large office with between 100 and 1,000 users: 

A typical user to trunk ratio is somewhere between 3 and 4, meaning that for up to 100 users a single T1 (23 trunk lines) or E1 (30 trunk lines) should be sufficient unless you run call center or conferencing applications.  Up to about 400 users are typically served with 4 T1 or E1 digital trunks. A single Audiocodes Mediant 1000 gateway supports up to 4 T1 or E1 trunk interfaces and additional analog FXS (station side for FAX or analog phones) or FXO analog trunk ports.

Hardware: Two 1U or 2U Servers (e.g. Dell, Intel or similar) with 4 GB RAM. Standard servers without any special HW or add-on cards. Two Audiocodes Mediant 1000 modular gateways (alternatively a Mediant 2000 or TP-260 PCI card can be used) Description: Supports mid-size to large offices with IP telephony incl. unified messaging, voicemail, auto-attendant, remote worker solution, desktop integration, FAX and conference phone support. Installs in hours. Grows with your business by adding gateways. Scales to multi-branch office deployments. Other gateways, analog or digital, can be added as needed where they are needed Redundancy:  This solution includes load-sharing redundancy for call control. It is easy to setup and provides high-availability for your phone system. In progress calls do not get interrupted if a server fails. Trunk fail-over and redundancy is offered between the two gateways. If one fails or all circuits are busy, the other takes over.

Deployment for large enterprise with more than 1,000 users (up to 5,000 or more): 

With deployments larger than about 1,000 users the sipXecs system is typically deployed in a distributed way where it's different components run on dedicated hardware that is centrally managed by the sipXecs configuration and management solution. Typical system partitioning would be as follows: 2 servers for redundant call control, 1 separate server for media services like voicemail and auto-attendant. The configuration management system would typically run on the same hardware as the media server or be installed on dedicated hardware. In addition, if larger call center capabilities are required, the ACD call center server that is part of sipXecs can also run on dedicated hardware. Several ACD servers can be run in parallel where different queues are assigned to different servers.

Distributed deployments where sipXecs components spread across several server hardware currently need to be manually installed and configured. Therefore, the installation process requires more technical skill and familiarity with the sipXecs solution to succeed. We are working on creating a cluster management system, where such deployments will become as easy as simpler installations. 

Performance: Since all the redundant sipXecs call control servers load-balance under normal operating conditions, call processing capability scales lineraly with the number of servers used. Two server redundant configurations are the best tested, but it is also possible to deploy 3, 4 or more servers in a redundant load-sharing cluster.  Media server performance is more critical for large deployments. The sipXecs media server, used mainly for voicemail, is memory bound and likes a lot of memory. Using 4 GB of RAM and a dedicated server should result in about 100 virtual media server ports. Increasing RAM to 8 GB will almost double that number. With sipXecs there is no limit on the total number of simultaneously active calls other than LAN bandwidth as calls are not routed through the sipXecs server. A nice side effect of this is better voice quality.

Deployment across several branch offices, small or large: 

sipXecs can be deployed as a multi-branch solution. There are two possible configurations to accomplish this:

1. Every office has its own installation of a sipXecs system. The different systems communicate among themselves to enable inter-office call routing based on a simple internal dialplan. Calls between offices can be routed over IP provided there is a VPN connection between the offices (no NAT). Alternatively, the offices can be connected over the PSTN.
   
2. A large centralized installation of sipXecs offers service to all the offces. Phones from the different offices connect over the corporate VPN (no NAT) to the central sipXecs server located in the data center. This setup provides complete feature transparency between all the offices. Local gateways can be located in remote offices to enable phones to make emergency calls even if the central sipXecs server or the WAN connectivity are unavailable. In addition, least cost call routing is possible deploying local gateways in remote offices, especially if these are global offices.

Support for FAX and analog Polycom (or other) conference room phones:

sipXecs supports both FAX and analog phones, such as existing conference room phones, using analog FXS gateways. FAX is received over analog or digital trunk lines, sent over IP to an FXS gateway, and out to a FAX machine.  sipXecs fully supports the new Polycom SoundStation IP4000 IP-based conference room phone.

Attendant console for receptionist:

There are several phones that support comfortable attendant consoles for sipXecs, among them Polycom, LG-Nortel and Snom. Also Grandstream has an attendant console. We are working on supporting Aastra phones, which come with a very nice LCD based attendant console. Presence indication (Busy Lamp Field - BLF) is available on the attendant console provided it is supported by the phone used. In addition the sipXecs Web user portal offers easy directory lookups with search for attendants to quickly dispatch calls. Speed dial entries can also be defined by individual users for their phones that can make the job easier for a receptionist. 

 Localization:

Starting with release 3.10 the sipXecs solution can be fully localized. Language packs can easily be uploaded and installed using the Web administration interface. Several languages can co-exist in the system giving every user the opportunity to select his or her desired language for the personal voicemail inbox. Voice prompts are localized, so is the language of the administration and individual user Web interface. in addition, there is a possibility to localize dialplans and call progress tones.  Phones can typically be localized as well, in many cases using the firmware management capabilities offered by sipXecs.

Installation:

Installation is easy. There is a single CD installation procedure that installs the complete operating system and the sipXecs application with no questions asked. A graphical wizard allows the administratr to give the system a hostname, IP address, timezone and a few other parameters. Phones and gateways on the LAN are auto-discovered and can easily be uploaded to the database. Users are configured with a few clicks using the Web admin user interface. Alternatively sipXecs can be synchronized with Microsoft Active Directory or you can upload user and phone configurations from a file created in Excel or similar spreadsheet application.

IT integration:

If you are a larger corporation with other IT infrastructure and you would like to integrate sipXecs into an existing system for fault and alarm monitoring or backup, then that can easily be done. In addition, sipXecs offers a Web Services SOAP based interface for all its configuration tasks which enables the easy creation of mashup Web portals either on the Intranet or external.

Integration with Microsoft Desktop, Active Directory and Exchange 2007:

We think integration with Microsoft is important. Using standard softphones from e.g. Counterpath sipXecs provides comprehensive integration with Microsoft desktoop applications. A softphone is available from Counterpath that integrates as a plugin into outlook and provides click-to-dial as well as video. Starting with release 3.10 sipXecs also provides GUI configured integration with Microsoft Active Directory, where users can easily be synchronized with a Microsoft environment. sipXecs also allows you to use the new Exchange 2007 voicemail and auto-attendant features where some users on your system can use Exchange VM and others can use the sipXecs voicemal system. All the necessary call routing is integrated into sipXecs and can be enabled with a simple click.

Who is using sipXecs?

sipXecs is used by probably thousands of companies. It is best known for its ease of use and rock solid stability. While there are many smaller companies using sipXecs, there are quite a few very large enterprises with several thousand, in a few cases over 5,000 users connected to a single redundant sipXecs system. We think you can trust the quality, stability, and performance of sipXecs. We also think there are significant benefits to choosing a system that is truly SIP standards compliant and interoperable. This is what makes your phone system future proof and does not lock you into a particular vendor's proprietary solution.