| [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] |
> I'm sure you must relalise that the reads are from memory and not disk! For the purposes of my test it doesn't really matter. I'm interested in the disk performance from the application level. If the application can read the data from disk very quickly it doesn't matter where it comes from. Tiobench does a pretty good job of performing consistent, repeatable tests. I received similar results when upping the file sizes for the tests. > Plus, to compare GFS and ext3 you must use the same locally mounted disk > for both tests - GFS used to support nolock for this sort of thing, > don't know if it still does. Yes, it does. lockproto=lock_nolock if my memory serves me correctly. Performance was _STILL_ faster than ext3. > What speed is the LAN over which you ran NFS? If it's Gigabit then > something is very wrong with your network! No, 10/100. And it is not dedicated just for testing so there was probably other traffic slowing things down a bit. Let me know if there are other tests you may be interested in? -Marc Swanson- ------------------------------------------------------- SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33 Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift. http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285 _______________________________________________ Opengfs-users mailing list Opengfs-users@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/opengfs-users