Once you have the system sizing recommendations, you can begin to order
the appropriate hardware configuration. First, however, you must determine how
many Unified ICM nodes you need.
The number of servers required in an Unified ICM system depends on the
configuration of the central controller, PGs, NICs, and other nodes. For
example, a duplexed central controller configuration requires additional
servers because the CallRouter and Logger are duplicated.
The following table shows how to determine the number of servers
required in your system.
The counts of servers in this example are based on an Unified ICM
configuration that has the following characteristics:
The Unified ICM system has a duplexed, geographically distributed
central controller (in other words, each central site has a CallRouter and a
Logger).
One side of the central controller (Central Site 1) is located at
a call center and consequently has a PG to serve one or more ACDs. The PG is
duplexed (two servers) for fault tolerance.
This Unified ICM installation has three remote call center sites
and two Admin sites.
Table 1 Sample server requirements
Sites
Node Types
CallRtr
Lgr
Call/Lgr
DB Server
PG
Administration & Data Server with HDS
Central Site 1
1
1
-
-
2
1
Central Site 2
1
1
-
-
-
1
Remote Call Center 1
-----
-----
-----
-----
2
-
Remote Call Center 2
-----
-----
-----
-----
2
-
Remote Call Center 3
-----
-----
-----
-----
2
-
Admin Site 1
-----
-----
-----
-----
-----
1
Admin Site 2
-----
-----
-----
-----
-----
1
Total Nodes:
2
2
8
4
Key:
--------- These servers are not installed at this type of
site.
– Not selected as an option in this particular configuration.
ICM platform considerations
Hardware & System Software Specification (Bill of Materials) for
Cisco Unified ICM/ContactCenter Enterprise & Hosted contains
information on server configurations and provides examples of supported server
platforms.
As a general rule for all Unified ICM nodes, keep processor
utilization below 60 percent of the maximum expected call load on the system.
This is needed to smooth out call request
"spikes" as well as to allow enough reserve capacity to perform
activities such as re-synchronization and background cleanup. Non-ICM software
can make up a part of the 60 percent maximum load. The processor utilization
figure (60 percent) covers all software running on the platform.
In addition to the utilization requirement, no software on the system
can run at a priority equal to or higher than the Unified ICM software for more
than 100 milliseconds in uninterrupted bursts. In other words, the Unified ICM
software needs to run on the system at least as frequently as once every 100
milliseconds. This is usually not a problem unless device drivers or other
kernel-level software is installed, or process/thread priorities have changed
incorrectly.
Note
Only installed at the central site if that site also serves as a call center or you are using the remote ACD option. A second central site is not required in duplexed-collocated Central Controller configurations.
Paging requirements
The most time-critical component of the Unified ICM system, the
CallRouter node, must not be delayed due to disk I/O (that is, paging). The
only disk I/O that should be occurring on Unified ICM machines is for log file
writes and database I/O. The database I/Os occur on Logger and Administration
& Data Server machines. The simple rule is to provide enough main memory so
that the entire working sets of critical processes remain in memory.
For complete and current information about RAM and other platform
requirements, see the
Hardware & System Software Specification (Bill of Materials)
for Cisco Unified ICM/ContactCenter Enterprise & Hosted.
Make sure the database platforms (Loggers, Administration & Data
Servers, and Unified ICM Gateway SQL machines) have enough main memory so that
all first level index pages are kept in main memory cache.
Logger expansion
The Logger platform you order can include a combination of internal
and external SCSI hard drives. As your call center enterprise grows, your
database requirements typically grow as well. You may have more services, skill
groups, and routes in your configuration, and you may be routing more calls
each day. This results in more historical data being stored in the central
database.
When your database requirements change, contact your Unified ICM
software support provider to have the storage capacity of the central database
increased.
Note
Refer to
Hardware & System Software Specification (Bill of Materials) for
Cisco Unified ICM/ContactCenter Enterprise & Hosted for more
information on data storage specifications.
You can allocate more database space after your system is installed
by:
Remotely adding database space (if current disk space allows).
Installing
"hot-plugable" disk drives and configuring the disks while the
system is running.
Note
Administration Guide for Cisco Unified ICM Enterprise
provides information for managing database space once the Unified ICM system is
installed and running.
Administration and Data Server planning
To allow users to monitor current call center activity, the Unified
ICM system forwards real-time data to Administration & Data Servers at
selected sites throughout the call center enterprise. The following figure
illustrates the real-time architecture of the Unified ICM system.
Figure 1. Real-time architecture of the Unified ICM system
Real-time call and agent group status data arrives at the central
controller from the Peripheral Gateways, which are constantly monitoring
activity at each call center. The CallRouter acts as the real-time server. The
CallRouter for the other side of the central controller acts as a back-up
real-time server.
The CallRouter is responsible for providing real-time data to one or
more Administration & Data Servers at each administrator site.
Administration Clients at the site receive their real-time data through a
connection to a Administration & Data Server. Administration Clients do not
have the local database and Administration & Data Server processes are
required to receive real-time data directly from the CallRouter.
Administrator sites
Administration & Data Servers can be located with one or both
sides of the central controller, at a call center, or at another site. An
administrator site is any site that contains Administration & Data Servers.
Each administrator site requires at least one Administration & Data Server.
You should use two Administration & Data Servers (as shown in
Administration and Data Server planning)
to provide fault tolerance in the real-time data distribution architecture.
The primary Administration & Data Server maintains an active
connection to the real-time server through which it receives real-time data.
The secondary Administration & Data Server also maintains connections to
the real-time server; however, these connections remain idle until needed (for
example, in cases where the primary Administration & Data Server is
unavailable for some reason). In sites that have two Administration & Data
Servers, the Administration Clients are configured to automatically switch to a
secondary Administration & Data Server if the first distributor becomes
non-functional for any reason.
Administration Client requirements
There is no set limit to how many Administration Clients can be served
by a Administration & Data Server. Refer to the
Hardware & System Software Specification (Bill of Materials) for
Cisco Unified ICM/ContactCenter Enterprise & Hosted for information
about requirements for Distributor and Administration Clients.
Historical Data Servers
Historical data is stored both as individual call detail records and
also rolled up and stored as interval records. An Administration & Data
Server with a Historical Data Server (HDS) stores historical data that supports
reporting queries. Administration & Data Servers at the site query
historical data from the HDS rather than directly from the Logger.
Figure 2. Historical Data Server Architecture
To set up a Historical Data Server, you must configure the Logger to
perform historical data replication. You must also configure the real-time
Administration & Data Server as an HDS. You can then create an HDS database
on the real-time distributor.
Information in the real-time feed tells each Administration Client
where to obtain historical data. If the real-time distributor is a Historical
Data Server, then it instructs its clients to get historical data from it.
Otherwise, it instructs its clients to get historical data from the Logger.
Each Logger can support up to two HDSs. The Administration Server, Real-time and Historical Data Server, and Detail Data Server (AW-HDS-DDS) can be enabled only on the primary distributor. The AW-HDS-DDS server role is disabled on the secondary real-time Administration & Data Server.
The HDS eliminates the performance impact on the central database
caused when multiple Administration & Data Servers need to access the
central database to generate reports. In systems that have multiple remote
Administration & Data Servers, the HDS brings Unified ICM historical
reporting data closer to the end user. Each HDS provides a set of database
tables. You can set specific times for retaining data in these tables. These
capabilities give you flexibility in setting up reporting capabilities on a
site-by-site basis.
The Historical Data Server also provides:
Greater flexibility in leveraging Internet applications.
An open interface for data mining and data warehousing
applications.
Host for other database tables and have them work with the HDS.
Improved security and data access capabilities.
The HDS Administration & Data Server requires a high-end
platform with a more powerful CPU, greater disk capacity, and more RAM. For
complete and current information on HDS requirements, see the
Hardware & System Software Specification (Bill of
Materials) for Cisco Unified ICM/ContactCenter Enterprise & Hosted.