I understand that Network Appliance has claimed that Oracle Dbase works better on a filer than a raw partition or even UFS, is this true?
I have been under the impression that Dbases work better on direct attached drives or even in a SAN.
I would personally verify that claim. IP based communication overhead introduces latency into the communication process. This is why even backup over a SAN is MUCH faster than over IP. There is no traversing an IP stack for communication and the Fibre Channel protocol latency is negligible. I would pit my 100MB connection to my Fibre disks against anyone's 100Mb IP connection to a filer anyday. Block based storageI/O is much more efficient than frame based File I/O. This is especially true in Oracle data mining applications. Raw throughput is needed here, and the SCSI protocol is more suited to this than IP. This is why the iSCSI folks are creating TOE (TCP/IP stack offload engines) in HARDWARE, so the stack is less in the way.
I would have liked to see the CPU utilization on the performance test they did. My bet is it was MUCH higher than when tested against a Fibre storage solution.
Dig deeper on Disk drives
Related Q&A from Christopher Poelker
SAN expert Chris Poelker discusses how to change the size of a LUN in a Microsoft cluster server environment.continue reading
SAN expert Chris Poelker compares connecting a SAN with wavelength cabling and dark fiber and discusses the pros and cons of each.continue reading
Storage expert Chris Poelker outlines WWN basics in order to answer the question: "Why do HBAs in a SAN have same base?"continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.