Hi, I would like to share my experiance as it was gained in post production where we beat our equipment to death!!!
Ok; a) determine site requirments b) determine budget c) establish install time d) predict hidden costs e) revise budget You are in a way, project managing this stuff so its very key that you get it right. What I found; Back when 10/100 was the shiznit, a couple of DAS served via Samba&NFS and distro'd the data for both load balacing and some form of redundancy. This distro'd both network and disk i/o. The network was the bottle neck here. Now, gig e is cheap and the burden has moved to disk i/o. Even when I distro amongst several 1TB servers, I am getting 50% network utilization while file access slows. This means that my Raids (DAS; Linux, SGI-XFS, Samba/NFS, discreeet IDE Raid subsystem to SCSI Ultra 160) are saturated i/o wise. To fix this, I must either redo my Raids so I have many more smaller disks (more spindles working) or get a NAS like a NetApp. I understand NetApp can go wire speed so this means that 50% network util means 50% disk util. Hope this helps you. Bri- PS The DAS approach is "a lot" more work in that I must maintain the OS, the IDE subsystem and the SCSI interface not to mention keep track of the distro'd data via Dfs for Win clients and autofs&symlinks for Unix clients. Look into NAS. The cost for either approach is about the same in terms of wether its all upfront or spread over time in terms of manitanance/headaches/etc... The advantage is peace of mind that a NAS buys you in that one can build a very robust DAS. __________________________________ Do you Yahoo!? Yahoo! Calendar - Free online calendar with sync to Outlook(TM). http://calendar.yahoo.com -- To unsubscribe from this list go to the following URL and read the instructions: http://lists.samba.org/mailman/listinfo/samba
