Dear readers, I read somewhere that linux raid1-mirroring will not improve the read performance of a single process but it will allow two separate processes to access the data at the same time. So when I access a 2-disk raid1-mirror with a single dd-process one drive will show disc-activity only. And if I kill the dd and restart it again, then the other disk will show activity. This lead me to the conclusion that a linux raid10-array with near-2 layout and an even number of drives will show the same behaviour. One single process will use one half of the drives. And a second process might use the other half in parallel. This seems to be wrong. A single dd on my 16-drive raid10-array will use all 16 drives and reads 704 MB/s. So here's my question: Does it make sense to create a raid10-array with near-2 layout and 2 drives only? Will such an array have twice the read-performance of a 2-drive raid1 mirror? Will a raid10-array with near-3 layout have 3 times the read performance of a single drive? Is there a way to create a raid10-array such that it will NOT use all drives but one half only? What about write-mostly for raid10-arrays? Kind regards Peter Koch -- 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
