Faq - Frequently Asked Questions
The following may serve as a guide to the DTraceToolkit.
16-May-2005, ver 0.30 (first version of the FAQ)
The DTraceToolkit is new, and as such there hasn't been many questions asked.
This may be better called a "possibly asked questions" :)
Questions
1. Intro
1.1. What is the DTraceToolkit?
1.2. Who wrote the DTraceToolkit?
1.3. Where do I get support?
1.4. Am I now a performance tuning expert?
1.5. Will this solve all my performance problems?
1.6. So the DTraceToolkit *is* DTrace?
2. Toolkit
2.1. What is in it?
2.2. What performance effect can the DTraceToolkit cause?
3. Contributing
3.1. Where do I send bugs?
Answers
1. Intro
1.1. What is the DTraceToolkit?
The DTraceToolkit is a collection of tools written using DTrace for
the Solaris 10[tm] OS by Sun Microsystems[tm]. Many of these scripts
will also work on OpenSolaris.
1.2. Who wrote the DTraceToolkit?
Volunteers of the DTrace and OpenSolaris community. Check the scripts
1.3. Where do I get support?
As the DTraceToolkit is a freeware product, there is no official company
offering support for this. Sun Microsystems does not support this. If you
post messages to the DTrace forums found in the Docs/Links file, a
volunteer may help you out.
1.4. Am I now a performance tuning expert?
The DTraceToolkit does not turn people into performance tuning experts in
the same way that owning a set of golf clubs won't make you a professional
golfer. Experience and understanding are necessary. The toolkit certainly
helps by fetching the data in an easy way, and also by providing some
documentation. So it is valuable, but not magical.
1.5. Will this solve all my performance problems?
This is similar to the previous point; the DTraceToolkit is valuable
for it's scripts and documentation, but it's no magical product.
Understanding and experience are necessary.
1.6. So the DTraceToolkit *is* DTrace?
The DTraceToolkit is one use of DTrace, but there is far more to DTrace
than just the toolkit. DTrace allows people to write their own customised
scripts to solve a wide number of problems.
Think of the DTraceToolkit as a starting point. Maybe your problem has
a solution in the kit. Maybe changing one of the toolkit programs slightly
is what you want. Finally you may need to write your script from scratch.
2. Toolkit
2.1. What is in it?
Read the Guide file for a table of contents, and Docs/Contents for a
list of commands.
2.2. What performance effect can the DTraceToolkit cause?
Enabling DTrace to monitor events has little effect on the system,
especially when compared to the disruptive behaviour of truss (See
http://www.brendangregg.com/DTrace/dtracevstruss.html for a comparison).
It really boils down to how often the events occur that you are monitoring.
The following numbers have been provided as an approximation:
1. Fixed rate scripts. For example, dispqlen.d samples at 1000 hz.
The impact will be negligible, close to 0% CPU. (in testing, 0.1% CPU).
2. Demand rated scripts. For example, iosnoop probes disk I/O events.
The impact depends on the rate of events, for many servers the disk
events would be slow enough for this to be less than 0.2% CPU.
Scripts such as execsnoop would expect even fewer events, their impact
would be close to 0.0% CPU. However scripts that monitor potentially
very rapid events will have a greater impact, for example running
dapptrace on Xorg (over 6000 lines of output per second) was consuming
around 10% of a CPU to do so.
3. Heavy voodoo scripts. A few scripts in the toolkit must probe either
a ton of different events, or very rapid events, or both. They are
going to hurt and there is no way around it. Scripts such as cputimes
and cpudists trace very frequent events, and can chew around 5% of
the CPUs; scripts such as dapptrace and dappprof trace extreamly
frequent events, and can chew over 20%.
There is an emphasis in the DTraceToolkit to write demand rated scripts
that measure the fewest events, such that their impact is close to 0.0%
CPU usage. Some scripts are fixed rate, which are safer as their impact
has a known upper bound, and are most suitable to run in production.
There are additional notes in Notes/ALLoverhead_notes.txt about the
overheads for running DTrace.
3. Contributing
3.1. Where do I send bugs?
The DTraceToolkit maintainer. See the Docs/Maintainer file.