Comparing System Monitoring Tools

I l @ ve RuBoard

This section provides a brief summary and comparison of the tools discussed in this chapter. The summary is organized by the key focus areas of system monitoring: configuration, faults, and resource and performance management.

Limited tools are available for monitoring configuration changes. The burden is largely on you to run tools such as iostat and STM to gather configuration data, and then to store that data in your own defined areas for comparison purposes later. The DMI Software MIF contains some software configuration information, but DMI management tools are not yet available. The Configuration Tracker is one tool that is available for monitoring software configuration changes. ClusterView can track changes made to high availability cluster configurations. More tools are needed that can track changes to hardware and software configuration information. This will be more important as it becomes easier to add or remove components while a system stays active.

System component failures can be reported through EMS for HP-UX, and SyMON for Sun Solaris systems. These products, with their automatic failure notifications, are preferred to using tools manually to probe the state of hardware components. Hardware monitors and monitor developer's kits are available for both products. Both can send notifications by using a variety of methods , including SNMP, but EMS has tighter integration with NNM and IT/O. SyMON provides integrated recovery actions with its product. Note that detecting failures of the system requires a remote monitoring tool, such as NNM.

If you don't have access to a performance management product, some low-budget performance monitoring can be done with UNIX commands, principally iostat, vmstat, top, uptime, sar, and swapinfo .

GlancePlus provides performance monitoring in real time for a large number of metrics on a single system. SyMON, by contrast, has very little performance data. For historical performance data, sampled over a longer time period, you should use a tool such as MeasureWare or BMC PATROL. If you want the tool to be integrated with a performance management product, then you may want to use MeasureWare, which provides tight integration with PerfView. Both products can report information via SNMP.

Some amount of resource monitoring can be done with performance monitoring tools or EMS. Additional resource usage data is available from ipcs, sysdef, and the HP-UNIX MIB.

I l @ ve RuBoard


UNIX Fault Management. A Guide for System Administrators
UNIX Fault Management: A Guide for System Administrators
ISBN: 013026525X
EAN: 2147483647
Year: 1999
Pages: 90

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net