Stanford Linear Accelerator Center

Winding Down Solaris Batch Service

SLAC Computing
UNIX at SLAC
Updated: 10 Aug 2005

SCCS currently runs a large batch cluster which includes both Solaris (Sparc) and Linux (Intel and AMD) worker machines. The Solaris systems, known as "broncos", are the oldest machines in the cluster and are reaching the end of their useful life. Although they have been quite reliable, they provide only a small fraction of the cluster's total computing capacity and they consume a disproportionate amount of power and cooling. We are therefore considering either shutting down entirely, or sharply reducing, our Solaris batch service sometime in the next couple of months. We expect the lost capacity to be more than offset by the next acquisition of new batch workers, which we plan to have in service by the end of October.

We would like to hear from any users who feel that completely shutting down this service would adversely impact their work. Please keep in mind that SCCS is not planning to eliminate all Solaris/Sparc resources. For example, we will continue to run the flora pool of public login machines and the tersk pool of interactive compute servers. For more information about the various categories of public UNIX machines and their intended uses, please see the Public Machines page.

Please send any comments about shutting down the Solaris batch service to unix-admin@slac.stanford.edu.

Len Moss
SCCS Systems Group


Len Moss