I am building a workstation for software development. The system will be compiling code for an average of 6 hours a day, every day. I would be curious to hear suggestions on the best disk arrangement for the system. The system will have 2.0GB of PC3200 RAM and running a 2.2GHz Athlon64. Yes, I know dual-Opterons would be ideal, but I am not made of money.
Anyway, I currently have the following disk arrangement in mind:
- 1 x 80GB SATA 8MB disc for the OSes (W2K and WXP-64), apps, and development environment
- 1 x 250GB SATA for the source code for the compile and general data storage.
But I am curious if I would see a gain from the following arrangement:
- 1 x 80GB SATA for the OSes and apps
- 1 x 80GB SATA partitioned for the development environment for each OS.
- 1 x 250GB SATA for the source code for the compile and general data storage.
My thinking with this arrangement is that it will limit wear on the HDDs and allow the highest throughput because each HDD has a different component for the compiles.
Now, I could RAID 0 the two 80GB HDDs but those arrays are prone to dying easily and I do not want to take the chance. Also, Raptors and SCSI are out of the question at the moment simply due to cost. For the 80GB HDDs I am thinking of Hitachi 7K80 7200rpm 80GB 8MB cache models and I already have the 250GB storage model.
Is there anyway to improve this arrangement for maximum throughput and fastest compile time? What else would you suggest?
Anyway, I currently have the following disk arrangement in mind:
- 1 x 80GB SATA 8MB disc for the OSes (W2K and WXP-64), apps, and development environment
- 1 x 250GB SATA for the source code for the compile and general data storage.
But I am curious if I would see a gain from the following arrangement:
- 1 x 80GB SATA for the OSes and apps
- 1 x 80GB SATA partitioned for the development environment for each OS.
- 1 x 250GB SATA for the source code for the compile and general data storage.
My thinking with this arrangement is that it will limit wear on the HDDs and allow the highest throughput because each HDD has a different component for the compiles.
Now, I could RAID 0 the two 80GB HDDs but those arrays are prone to dying easily and I do not want to take the chance. Also, Raptors and SCSI are out of the question at the moment simply due to cost. For the 80GB HDDs I am thinking of Hitachi 7K80 7200rpm 80GB 8MB cache models and I already have the 250GB storage model.
Is there anyway to improve this arrangement for maximum throughput and fastest compile time? What else would you suggest?