Using distcc to speed compilation

Michael mutk at
Sun Nov 17 19:19:36 PST 2002

Hi all,

I have been experimenting with distcc ( to 
help speed up compilation of large applications..

So far I have built only a Kernel, and with great success.

I have a P4 1.9Ghz system, running Redhat 8.0, and VMware. I have 
installed a LFS system on a VMware Virtual Machine. I have compiled 
and installed distcc on the LFS system , and it works straight out of 
the box with no build options (It installed binaries in /usr/local/bin).

I also have another system, running Redhat Linux 8.0 and distcc 
(Installed from RPM) . This system is a Celeron 850Mhz Coppermine CPU.

To illustrate the speed up, I installed GNU 'time' tool on the LFS 
system and present the time taken to do a Kernel compilation on the 
VMware VM running LFS.

## Without distcc , kernel build on VMware LFS system running on P4 
1.9Ghz CPU ##
- do the usual
	make mrproper && make oldconfig && make dep
- do
	time 'make bzImage'
- result:
real 6m18.823s
user 3m40.200s
sys 2m33.570s

## Using distcc. An external host (Celeron 850Mhz) running distccd to 
share CPU cycles for the Kernel compilation ##

- result:
real 4m35.220s
user 2m34.940s
sys 1m57.000s

If the system I was stealing CPU cycles from was faster, it would 
approach half the normal compilation time. I have done a Kernel 
compile on a Celeron 850Mhz system, using the P4 1.9Ghz system as 
slave, and got better than half the usual time.
I think distcc deserves a mention in the BLFS somehere, perhaps as a 

I know that building XFree86 is a lengthy process so being able to 
shave time off it would be a distinct advantage...

Unsubscribe: send email to listar at
and put 'unsubscribe blfs-support' in the subject header of the message

More information about the blfs-support mailing list