Sizing the Server for the Personify Database

The following are two primary measurements that should be used when sizing a server for the Personify database:

1.    Total concurrent users: This is the number of people that will have the Personify back office application open on their desktop during the peak times of the day. Break it down by data entry users (both power data entry and casual data entry), view only users, reporting users, and dashboard users.

2.    Total number of system records:  This gives you an idea what kinds of data we will be maintaining in the system.

a.    Check with customers what they are license for. 

b.    How many customers are being created each day?

c.    How many orders created each day (not just order master records but line items on each order)?

d.    How many receipts?

e.    How many contact tracking and activity records?

f.      How many notification emails get sent each day? This is a total number of these items regardless of whether they are created by staff or from the web through the self-service portal

g.    How many and what types of reports does the customer anticipate running per day?  

The table below provides a reference point for sizing based upon the above criteria.  The sizing is referenced by one of six different classifications: x-small, small, medium, large, x-large, and enterprise.  Configuration for each one of these server classes is located below the table.

Server Classifications:

1.    X-Small

a.    Processors: One socket, 4 core or 4 virtual processors

b.    Memory: 8 GB

c.    Disks: Raid 5 or SAN equivalent, SATA 7200 RPM, Minimum 3 disks, Single volume for all database files

2.    Small

a.    Processors: One socket, 6 core or 6 virtual processors

b.    Memory: 16 GB

c.    Disks: Raid 5 or SAN equivalent , SATA 7200 RPM, Minimum 5 disks, Single volume for all database files

3.    Medium

a.    Processors: Two socket, 6 core or 12 virtual processors

b.    Memory: 32 GB

c.    Disks: Raid 5 or SAN equivalent, SAS 10K RPM, Minimum 5 disks, Separate volumes for database data files, transaction logs, tempdb data and backups. 

4.    Large

a.    Processors: Two socket, 8 core or 16 virtual processors

b.    Memory: 64 GB

c.    Disks: Raid 10 or SAN equivalent, SAS 10K RPM, Minimum 5 disks, Separate volumes for database data files, transaction logs, tempdb data and backups. 

5.    X-Large

a.    Processors: Four socket, 6 core or 24 virtual processors

b.    Memory: 128 GB

c.    Disks: Raid 10 or SAN equivalent, SAS 15K RPM, Minimum 6 disks, Separate volumes for database data files, transaction logs, tempdb data and backups.  Use of partitioning technology.

6.    Enterprise

a.    Processors: Four socket, 8 core or 32 virtual processors

b.    Memory: 258 GB

c.    Disks: Raid 10 or SAN equivalent, SAS 15K RPM, Minimum 8 disks, Separate volumes for database data files, transaction logs, tempdb data and backups.  Use of partitioning technology.

Other sizing criteria should be considered before making the final determination on database server sizing:

1.    Total number of Personify back office users: This is simply a total count of the number of Personify back office users that will have access to the application.  This includes all users, even the most occasional user. 

2.    If using Personify e-Business,

a.    Total expected visitors per day during peak periods: This should simply be the number of expected unique visitors to e-business during peak periods. Anonymous or registered user?

b.    Total transactions expected a day during peak periods: A transaction is any interaction with the e-business website.  This could be a purchase for their store, update of customer data or membership renewals. 

c.    Type of reports: How many and what types of reports does the customer anticipate running per day?  

3.    Customizations to the system. (i.e., new user defined columns / tables) How much data is going to be stored in these new user defined columns / tables. Is this data read only or will it be updated regularly? Is this data that is being stored that is not accounted for in base Personify?

4.    Will the customer be implementing a data warehouse (Business Advantage) and running ETL’s: Yes or no. This helps determine whether or not ETL’s will be run to extract data from production databases and into a data warehouse. Any large import processes can impact the system. Is the customer importing data on a regular basis (daily, weekly, monthly)? Determine what volume of data that would be.

5.    External Systems.  Any external systems that will contribute to the load on the Personify database?

6.    Auditing Solution.  Will you implement the audit logging solution, Apex? If so, what tables / columns are they auditing?