Home
Tuning
Use this information to improve HCL Domino® server, Domino Web server, and messaging performance through the use of resource balancing and activity trends, Server.Load commands, advanced database properties, cluster statistics, and the Server Health Monitor.
The Server.Load tool
Server.Load is a capacity-planning tool that you use to run tests, also called "scripts" and "workloads," against a targeted Domino® server to measure server capacity and response metrics.
Server.Load metrics and messaging statistics
As you run a test, you can view various script metrics and server statistic metrics and optionally store the test output in a separate file. Server statistic metrics are generated by the Domino® server. Script metrics correspond to Server.Load command names and display the performance of particular commands. For example, if you select the Add metric, the Metrics window displays the results of the Add command.
HTTP messaging statistics
When an HTTP workload is run in Server.Load, an HTTP messaging statistics window displays (in addition to the metrics window).

Tuning
Use this information to improve HCL Domino® server, Domino Web server, and messaging performance through the use of resource balancing and activity trends, Server.Load commands, advanced database properties, cluster statistics, and the Server Health Monitor.
- Resource balancing and activity trends
  Domino® server resource utilization can be separated into two types, system activity and user activity. System activity, which includes the level of processor, disk, memory, and network consumption that Domino generates to keep the server running, is a fixed amount of activity, as long as systems are healthy and performing smoothly. Domino servers typically use a modest percentage of their resources to run. The remaining server capacity is used to support user activity, which varies with the usefulness of the data on the server.
- The Server.Load tool
  Server.Load is a capacity-planning tool that you use to run tests, also called "scripts" and "workloads," against a targeted Domino® server to measure server capacity and response metrics.
  - Server.Load agents
    Server.Load includes a set of agents in the file NAMAGENT.NSF, which is initially installed in the data directory on the Domino® Administrator client. The first agent in this list -- Create NotesBench Mail Person Documents -- is used to set up Person documents for the workloads and set the HTTP password. The rest of the agents are used to repair and change the workload setup.
  - Server.Load metrics and messaging statistics
    As you run a test, you can view various script metrics and server statistic metrics and optionally store the test output in a separate file. Server statistic metrics are generated by the Domino® server. Script metrics correspond to Server.Load command names and display the performance of particular commands. For example, if you select the Add metric, the Metrics window displays the results of the Add command.
    - HTTP messaging statistics
      When an HTTP workload is run in Server.Load, an HTTP messaging statistics window displays (in addition to the metrics window).
    - Monitoring Server.Load metrics
      You can monitor various script metrics and server statistic metrics.
  - Setting up clients and servers for Server.Load
    To use Server.Load, install the Domino® server on the server under test (SUT) and install the Domino Administration client and the Server Load Utility on each client.
  - Built-in and custom Server.Load scripts
    Server.Load includes a set of built-in scripts. You can also create a custom script from scratch.
  - Running the built-in Server.Load workloads
    Server.Load contains numerous built-in workloads and initialization workloads.
  - Server.Load commands
    Server.Load scripts consist of statements in a simple command language, the Server.Load specification language.
  - Server.Load scripts
    You can use any of the sample Server.Load scripts as they are or you can use them as a basis for creating customized scripts.
  - How Server.Load determines the range of mail addresses it uses when sending messages
    Server.Load uses the logic described in this topic when determining the range of mailing address to be used.
- Tuning Domain Indexer performance
  Each time the Domain Indexer task runs, it looks in the Domain Catalog for new databases that have the Include in multi database indexing property enabled. It then looks for documents and files in existing databases and file systems that are new or changed since the last time it ran, and adds them to the Domain Index.
- Improving Domino® server performance
  Read the following topics for help on improving basic Domino® server performance and capacity, as well as the performance of these features: Agent Manager, databases and the Domino directory, the directory catalog, LDAP searches, mail, Web server, and UNIX™ server.
- Improving Web server performance
  After you set up the Domino® Web server and make sure that it runs properly, check the server's performance and response time.
- Improving mail performance
  Domino® includes features that improve efficiency in specific environments, but these features may not be switched on by default.
- Setting advanced database properties
  Advanced database properties include performance optimization and compression features, as well as ways to manage usability features such as unread marks and soft deletions.
- Understanding cluster statistics
  Three categories of Domino® cluster statistics help you analyze clusters
- Improving the performance of the Server Health Monitor
  If the Domino® Administration client workstation performs at 100 percent CPU utilization for a long period of time, the Server Health Monitor discards server statistic data to keep up with the workload.

HTTP messaging statistics

When an HTTP workload is run in Server.Load, an HTTP messaging statistics window displays (in addition to the metrics window).

Table 1. Messaging statistics that display in Server.Load
Messaging Statistic	Description
ConnectTime	Minimum, maximum, and average number of milliseconds required to establish an HTTP connection to the server.
SendTime	Minimum, maximum, and average number of milliseconds to send HTTP data to the server.
RecvTime	Minimum, maximum, and average number of milliseconds to receive HTTP data from the server.
TotalTime	Total minimum, maximum, and average number of milliseconds to connect to the server and to send and receive HTTP messages.
ActionsTime	Minimum, maximum, average number of milliseconds to perform an HTTP action. An HTTP action is an HTTP request that involves special work on the server to process. For example, in the case of the iNotes® workloads, an action is a POST request that involves sending or deleting mail.
Kbytes/Page	Minimum, maximum, and average number of kilobytes of data received per request.
Object/Page	Minimum, maximum, and average number of objects (htm, jpg) of data received per request.
Totals	Kbytes = Total number of kilobytes of data received Pages = Total number of Web pages received Obj = Total number of Objects received Timeout = Total number of server timeouts, number of times the server was not responding Drops = Total number of times the server dropped the connection, the number of times unable to send or receive after successfully establishing an endpoint ActF = Total number of actions that fails, number of times unable to connect, send, or receive after successfully establishing an endpoint for an Action request.
Act	Total number of actions
Serr	Total number of server errors
Cerr	Total number of client errors
Hit/Min	Average number of objects received per minute
Min Running	Number of minutes that the test has been running