Tivoli Monitoring Tool Configuration Task
· Help identify what metrics out of the separate requirement document can be implemented with IBM Tivoli Monitoring tool.
· Configure ITM tool (ITCAM for Web Response Time, for OS and for SQL Server) to monitor on system metrics as defined in the separate requirement document.
· Resolve the performance issue with Tivoli tool (ITCAM WRT is very slow with about 0.5 million transaction records).
· ITM tool keep 24 hours of real-time transaction instances. For historic (over 24 hours) record, the tool takes data sampling instead of averaging – this causes loss of true transaction tracks. We need to find a solution to capture the true historical performance data.
AssistPlus Monitoring Requirements for IBM Tivoli Tools.
1 Overview.
This document describes the requirements for AssistPlus monitoring using IBM Tivoli tools. The monitoring is carried out by the following ITM Agents:
Operating System (OS) Agents installed on the Web and database servers;
Web Response Time (WRT) Agent installed on the Web servers;
SQL Server Agent installed on the database servers;
Robotic Monitoring Agent installed on selected client machines.
The requirements in this document do not cover monitoring of DocProd, Interact, or MSMQ servers, since no Tivoli Agents are currently installed on these servers. Also, the requirements do not include operational server-level (rather than application-level) monitoring such as monitoring of the server temperature or the amount of free disk space.
We assume that the OS Agent is capable of capturing all or most of the system-level performance metrics available through Windows Performance Counters, while the SQL Server Agent is capable of capturing all or most of the database-related Performance Counters. Therefore, the requirements discussed in this document are based on the performance metrics available through the Windows and SQL Server Performance Counters. Lower-level SQL Server monitoring requirements, such as tracing performance problems to specific SQL statements, are outside of the scope of this document.
Building a practically useful set of performance metrics is an incremental process. Therefore, the set of the performance metrics described in this document is by no means exhaustive, and additional metrics will be added to this set in the future.
2 General Requirements.
1. All the performance metrics discussed below (with the exception of synthetic transaction metrics) should be collected continuously throughout the day and averaged over successive 5-min intervals. A chart that plots the daily values for each metric as a function of time over successive 5-min intervals must be provided. A user should be able to view intraday data for any selected day.
The synthetic transaction metrics should be collected at 5 min intervals and displayed in intraday charts similar to other metrics.
2. In addition to the intraday chart described above, a daily chart for each metric should be provided as well. The daily chart displays an average value of the metric between 8 am and 6 pm on each day over the time period selected by a user. The readings taken outside of the 8 am – 6 pm interval (in particular, the readings pertaining to the overnight process) should be excluded from the averaging.
3. Intraday data should be retained for at least 2 months. Daily data should be retained for at least 6 months.
0 comments:
Post a Comment