This web log will is the start of a small series of Solaris performance monitoring tools. We are going to present freely available tools from Solaris, the Tools CD and the DTrace toolkit.
This blog will help the upatient ones who want to solve 90% of the day to day performance issues without long studies.
Being able to scale quickly is pivotal in the web age where the load can grow by an order of magnitude overnight. Our little primer will release smaller chapters which are going to apply the following methodology:
- Chapter 2: Document your performance issue: How to instrument my data center with dimSTAT to get an idea when the performance worries happened
- Chapter 3:What are the features and limits of my system: How to analyze your Solaris installation, patches etc.. Quantify the abilities of your hardware: How many processors, disks, memory etc.do I have?
- Chapter 4: What is my system actually doing? Checking out your processes, top users, resource consumption. Understand your system utilization.
- Chapter 5: What is my process doing? Tools for process introspection. Learn which files are being used, which libraries are being loaded, which call stacks you're currebtly using etc.
- Chapter 6: Monitoring the processes and threads in projects and zones with prstat
- Chapter 7:Understand your IO: Who is writing how to where? How the the disk sub system doing?
- Chapter 8: Understanding network traffic: Which network interface is working how hard? Is my interface overloaded etc.
- Chapter 9: Tracing at large: How to get information about current system call. How to use DTrace to answer common questions like: Who is writing to which file right now?
This upcoming sequence of blogs will hopefully answer the most common performance questions with standard tools available for Solaris.
Other sites which have Solaris Performance related information are:
Posted by Darryl Gove's blog on May 01, 2008 at 06:36 PM CEST #