Solaris Page Update: How To Use DTrace
The Network Management Solaris page has been updated, adding a reference to DTrace.
There is a feature embedded into Solaris, starting with Solaris 10, which allows for systems administrators to get telemetry (i.e. live data & statistics) from a running system with virtually no overhead. This is accomplished by the instrumentation of the OS (i.e. Solaris) at the kernel level, with something that is called "probe points". Any common systems administrator can diagnose a third party application performance by observing the Solaris "probe points" and measuring performance via DTrace.
Does it still sound strange?
It very well could, unless you have an operating system which is BSD derived. Apple OSX and Solaris are two operating systems which leverage this free infrastructure. Network Management resources should be skilled in this feature, to understand the scalability of their applications, operating systems, and hardware platforms.
Solaris Reference Material
- DTrace.org - Blog from a creator of DTrace
- Solaris Dynamic Tracing Guide
- How To Use DTrace
- Introduction to the DTrace Book [Video]
- What is in the DTrace Book [Video]
- DTrace Book Released! (Book Samples: Table of Contents, File System Chapter)
- Observing File System Latency with MySQL using DTrace: Part [1|2|3|4]
- DTrace with Oracle VM Server for SPARC 2.1
Solaris LDoms / Oracle VM Server SPARC