DCC build options for performance and sizing

Gary Mills mills@cc.umanitoba.ca
Fri Jun 11 00:54:20 UTC 2010


On Thu, Jun 10, 2010 at 01:46:18AM +0000, Vernon Schryver wrote:
> 
> Judging from their graphs on the server status web page, I suspect only
> one of the two umanitoba.ca servers is running a 64-bit dccd, and that
> the larger pointers let dccd take advantage of the more than 4 GByte
> of RAM on the system.  That forced the dccd buffer page or chunk size
> to be twice as large or even larger.  The active hash table entries for
> the same number of flooded or local transactions are spread among twice
> as many bytes.  Either or both effects caused twice as many bytes to
> be flushed to the dcc_db.hash file.

You're talking about dcc2, I presume.  On Saturday 5 June, I migrated
it from a Solaris 9 SPARC server with 2 GB to a Solaris 10 x86 server
with 32 GB.  It started with an empty database at that time.  I see
that the checksum count has been ramping up since then.  Last night,
I reconfigured it from write() to tmpfs for the hash files.

dcc1 has not changed since I upgraded it to dcc-1.3.125.  Both are
running the same daemon:

    $ file /usr/local/dcc/libexec/dccd
    /usr/local/dcc/libexec/dccd:    ELF 64-bit LSB executable AMD64 Version 1 [SSE2 SSE FXSR FPU], dynamically linked, not stripped

-- 
-Gary Mills-        -Unix Group-        -Computer and Network Services-



More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.