POM system capacity

Last Updated : Oct 31, 2025 |

This section lists POM system function limits. The limits described are based on performance test results in a lab. The performance of a production POM system may vary due to factors beyond product code such as deployment, environment, and integrations.

POM is dependent on multiple environments and solution elements for the peak performance of Busy Hour Call Attempts (BHCA) and other list and campaign management operations. The following are some factors that can influence the performance of POM operations:

  • POM deployment

  • POM system performance (CPU, memory, Disk I/O) – bare-metal hardware or VMWare

  • Network latency between POM components deployed on multiple hosts (multi-POM setup)

  • Network latency between POM and other solution components such as MPP, Application server, Database, and Widget server

  • Database response time

  • MPP response time

  • Application Server response time

  • Customizations

Therefore, when you review the performance of production operations, be aware of these factors to ensure you meet the pre-requisite operating performance level for POM to function normally and optimally.

The following table displays the maximum operating limit of POM Campaign and Agent functions:

Description

Capacity

Notes

Maximum BHCA per POM system

Up to 500,000

For more information on the tuning parameters and tested scenarios, see Tunable parameters for performance and example deployment scenarios.

Maximum total contact records that can be imported per hour from the file data source.

500,000

Import speed may reduce if advanced options such as DNC check, Duplicate contact - update existing check, Automatic update of time zone, are enabled.

An increase in the number of parallel jobs can also impact import speed.

Maximum total contact records that can be imported per hour from database data source.

400,000

Import speed may reduce if advanced options such as DNC check, Duplicate contact - update existing check, Automatic update of time zone, are enabled.

Maximum number of simultaneous import job

30

The number of maximum simultaneous import job depends on the hibernate.hikari.maximumPoolSize parameter defined in the PIMHibernate.cfg.xml file. The number of maximum simultaneous import job will be 30% of the value defined for hibernate.hikari.maximumPoolSize parameter. The default value for hibernate.hikari.maximumPoolSize parameter is 100. So default value for the number of maximum simultaneous import job is 30.

Note:

To ensure optimum performance and stability of POM, restrictions are imposed on number of parallel imports supported as well as on the number of contacts that can be imported or uploaded into the contact list at a time when POM is under high load. For more information on restrictions on imports, see Restrictions on imports or uploads when POM is under high load.

Maximum number of agents

2,000

The maximum number of agents supported per zone is 1000. Therefore, a minimum of two zones is required for 2000 agents.

Maximum number of Rules for the system

50

-

Maximum number of Rules that can be associated per campaign

10

Each rule that you associate with a campaign can impact the performance.

Factors such as type of rules, validation conditions in the rule, and complexity of the database queries can impact the run-time performance, that is, time taken for the rule evaluation. This in turn can impact POM dialing performance.

Maximum number of Records per contact list

200,000

Number of records in a Contact list can impact speed of import, speed of filtering, and speed of dialing. Hence, Avaya recommends the use of a contact list no longer than the size of 200,000.

Maximum number of contacts per campaign

200,000

Customer must ensure that the total number of contacts getting filtered for each campaign always remains below the maximum supported value. Otherwise, there can be a major impact on system performance.

Maximum number of contact attempts in the database

30,000,000

Customer must configure the POM purge schedule such that the total number of records in the pim_contact_attempts_history table in the POM database always remain below the maximum supported value. Otherwise, there can be major impact on system performance.

POM does not monitor growth to this database table. Customer DBA must monitor and determine average daily growth rates and configure retention period such that the number of records in this table does not grow beyond the supported limit. Growth of this table beyond the supported limit can impact POM dialing performance. Customer DBA must perform regular database maintenance and reindex for optimal performance.

Maximum number of Records in contact table

5,000,000

Number of records in a contact table can impact the speed of import, speed of filtering, speed of dialing, and speed of the Show all contact feature on the contact list page. Hence, Avaya recommends cleaning up the contact table if it has reached the maximum capacity.

The following table displays the maximum operating limit of the Contact parameters of POM:

Parameter

Maximum limit

Attributes that can be created in the POM system

900 (including the system attributes)

Attributes in a contact list

150

Phone numbers in a contact list

10

Attributes in a contact import file on any contact import data source

150

File size allowed to upload using the web interface

50 MB

File size for SFTP/Local file-based data source import

1 GB

The following table displays the maximum operating limit of the DNC parameters of POM:

Parameter

Maximum limit

DNC lists in system

200 DNC lists allowed for system

DNC groups in system

200 DNC groups allowed for system

DNC groups per campaign

10 DNC groups are allowed during campaign creation

DNC lists per DNC group

10 DNC lists are allowed per DNC group

The following table displays the maximum operating limit of the Campaign parameters of POM:

Parameter

Maximum limit

Simultaneous campaign jobs

200 (if using only 1 handler per Campaign job)

For a multi-handler campaign, each handler is counted as a separate campaign for this limit.

Agents

2000 per POM setup across multiple zones

Maximum 1000 per zone

Agents per Agent Manager

1000

Only 1 Agent Manager per zone

Number of scheduled activities per minute, such as import job and campaign job

100

Supervisor: Maximum agent groups

500

Supervisor: Maximum agent per group

500

Email attachment size

25 MB

Scheduled jobs per minute such as import job, campaign job, and so on.

100

Schedules per campaign

50

Contact list import without load

You can have a maximum of 900 attributes (including system and custom attributes) in the POM system.

For the following tests, the number of custom attributes created in the system is 150. The test was performed by importing files with a varying number of records and a constant record size from the local file data sources. Each record had attributes of type Integer, Long, Short, Char, Float, Boolean, String, Phone, Email, Date, Time, and Timestamp. The size of each record was 1.296 KB, with 100 attributes per record, including 13 system attributes, 5 phones attributes, and 32 other custom attributes. The default contact import batch size was used.

The Contact Import was performed using the parallel import feature. The feature can be enabled by clearing the Maintain insertion order during contact import check box on the Global Configuration screen. Also, the feature to store the import status of successful records in the POM database was disabled by clearing the Record successful import check box on the Global Configuration screen. Log level was configured with Info level.

When the Contact import runs, no other campaign or export runs on the POM server.

The following table depicts the different import tests performed:

Table 1: Time required for contact list import without load

Import Run

Number of Records

Total Attributes (system + Custom)

Approximate Time Required for import

Run 1

10,000

100

1 minute and 2 seconds

Run 2

50,000

100

4 minutes and 58 seconds

Run 3

100,000

100

10 minutes

Run 4

500,000

100

20 minutes and 32 seconds

The data in the table is for a file-based import from a local data source when the POM database is on an external MS SQL server. The POM server and POM database are configured on Model – Dell Inc. VxFlex-R640, 40-GB memory, and 28 CPU – Intel(R) Xeon(R) Gold 6248 CPU @ 3.00 GHz with 10-GBPS network speed.

For testing the data source in POM, select the following advanced import options:

  • Automatically update time zone for phone numbers

  • Check phone numbers for reject patterns

  • Check phone numbers for phone formats rule

You may experience more time for import if you select additional options for the data source.

Note:

You can configure multiple data sources for the same contact list and run those concurrently to improve the import speed. A contact list importing 10,000 records using a single data source takes more time than the same contact list importing the same number of records divided into three or more data sources.

To import 10,000 records in a contact list, you can configure 3 data sources. Each data source fetches distinct records.

Datasource1 imports the first 3000 records. Similarly, datasource2 imports the next 3000 records, and datasource3 imports the remaining 4000 records.

Recommendations

  • The records are not sorted in the POM database. Using filter and sort conditions in campaigns can address this issue.

  • Do not select the Empty Contact List before Import option for these data sources. You might need to empty the contact list manually before starting the multiple data sources.

  • Divide the records between 3 to 6 data sources.

Contact list import with load

A load of 200 jobs and 1000 agents runs while the contact list import test is performed. The POM server is configured on Model – Dell Inc. VxFlex-R640 with 40-GB memory and 24 CPU. The MS SQL server is configured with 28 CPU and 40-GB RAM – Intel(R) Xeon(R) Gold 6248 CPU @ 3.00 GHz with 10-GBPS network speed. The Finest log level is enabled.

The following table depicts the different import tests performed:

Table 2: Time required for contact list import with load

Import run

Number of records

Total attributes (System + Custom)

Approximate time required for import

Load Monitor service status

Run 1

10,000

100

4 minutes and 38 seconds

Enabled

Run 2

10,000

100

4 minutes and 25 seconds

Disabled

Run 3

50,000

100

19 minutes and 5 seconds

Disabled

Run 4

100,000

100

36 minutes and 5 seconds

Disabled
Note:

The new Load Monitor service delivered in POM 4.1 implements guard rails that prevent loading many contacts with traffic.

For testing the data source in POM, select the following advanced import options:

  • Automatically update time zone for phone numbers

  • Check phone numbers for reject patterns

  • Check phone numbers for phone formats rule

You may experience more time for import if any or all such options are selected for the data source.

Bulk contact list import using REST API

This is a new feature in POM 4.0.2 which allows a customer to import contact list data using REST API. Customer data must be mentioned in JSON body format.

The test was performed by importing files with a varying number of records and a constant record size from the local file data sources. Each record had attributes of type Integer, Long, Short, Char, Float, Boolean, String, Phone, Email, Date, Time, and Timestamp. The size of each record was 1.296 KB, with 26 attributes per record, including 8 system attributes, 2 phone attributes, and 10 other custom attributes. The default contact import batch size was used.

The Bulk Contact Import by REST API was performed using the parallel import feature. The feature can be enabled by clearing the Maintain insertion order during contact import check box on the Global Configuration screen. Also, the feature to store the import status of successful records in the POM database was disabled by clearing the Record successful import check box on the Global Configuration screen. Log level was configured with Info level.

When the Bulk Contact import runs, no other campaign or export runs on the POM server.

The following table depicts the different import tests performed:

Table 3: Bulk contact list import using REST API

Import run

Number of records

Total attributes (System + Custom)

Approximate time required for import

Run 1

10,000

20

18 seconds

Run 2

50,000

20

1 minute and 27 seconds

Run 3

100,000

20

3 minutes and 9 seconds

The data in the table is for a file-based import from a local data source when the POM database is on an external MS SQL server. The POM server and POM database are configured on Model Dell Inc. VxFlex-R640, 40-GB memory, and 28 CPU – Intel(R) Xeon(R) Gold 6248 CPU @ 3.00 GHz with 10-GBPS network speed.

For testing the data source in POM, select the following advanced import options:

  • Automatically update time zone for phone numbers

  • Check phone numbers for reject patterns

  • Check phone numbers for phone formats rule

You may experience more time for import if you select additional options for the data source.

Bulk contact import with load

A load of 200 jobs and 1000 agents runs while the contact list import test is performed. The POM server is configured on Model – Dell Inc. VxFlex-R640 with 40-GB memory and 24 CPU. The MS SQL server is configured with 28 CPU and 40-GB RAM – Intel(R) Xeon(R) Gold 6248 CPU @ 3.00 GHz with 10- GBPS network speed. The Finest log level is enabled.

The following table depicts the different import tests performed:

Table 4: Bulk contact list import with load

Import run

Number of records

Total attributes (System + Custom)

Approximate time required for import

Run 1

10,000

20

30 seconds

Run 2

50,000

20

2 minutes and 28 seconds

Run 3

100,000

20

5 minutes and 18 seconds

For testing the data source in POM, select the following advanced import options:

  • Automatically update time zone for phone numbers

  • Check phone numbers for reject patterns

  • Check phone numbers for phone formats rule

You may experience more time for import if any or all such options are selected for the data source.