On Fri, 5 Dec 2003 09:43:13 -0800 Bob Miller <[EMAIL PROTECTED]> wrote: | > On Thu, 04 Dec 2003 19:26:51 -0600 Steven Elling <[EMAIL PROTECTED]> | > wrote: | > | P.S. Developers: You might want to look into adding a number to | > | the'-j' option in MAKEOPTS during emerges of certain packages to | > | account for this situation. | > | > Wouldn't that be rather unfair upon users who have a 100 CPU distcc | > cluster? | | My testing showed that distcc doesn't seem to scale beyond about 8-10 | hosts. (Hosts, not CPUs.) Since the originator has to do all the cpp | processing and all the linking, the other hosts start standing around | idle when the originator reaches 100% CPU.
It depends a lot upon what the central host is. If the central host is faster and has several CPUs (for example, a server using workgroup machines as slaves) then distcc can potentially scale very well. If the central host is a really old sparcstation5 then adding more than about two slaves doesn't help. It also depends upon the makefile in question. Very few Makefiles allow more than maybe 20 things to be compiled simultaneously anyway. So maybe I should rephrase... Wouldn't that be rather unfair upon users who have a dozen 8-way amd64 boxes fully interconnected by switched gigE who do glibc development? -- Ciaran McCreesh Mail: ciaranm at gentoo.org Web: http://dev.gentoo.org/~ciaranm
pgp00000.pgp
Description: PGP signature
