> I'm very sorry but now i can't confirm anymore that stripe_cache_size helps. > > My test was to short. With every minute the IOPS decrease. So > stripe_cache_size does only help for very short tests. > > I will provide details in another post. > Now i wish i could delete the post. As the SSD performance decrease with running time (within minutes) i'm not sure if the better performance with higher stripe_cache_size would be constant over time or is more a effect from a fresh empty cache that ceases with time (in my scenario with writing many small random blocks) Short tests results: Raid 5, 3 SSD stripe_cache_size noop deadline (tuned) 256 18914 18730 16384 18161 19766 Raid 5, 4 SSD stripe_cache_size noop deadline (tuned) 256 11863 13716 16384 13186 14688 Peter -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
