Hi. We have a busy mailserver with the mail-spool on MD (raid1, consisting of sda and sdb, ext3 w/data=journal noatime nodiratime, 2.6.6, Debian Woody). The read-balancing does not seem to work properly. Output from "iostat -x /dev/sda /dev/sdb 1" (edited for readability) shows the following: Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s rkB/s wkB/s avgrq-sz avgqu-sz await svctm %util sda 4.00 1.00 217.00 8.00 2472.00 88.00 1236.00 44.00 11.38 147.22 706.14 4.45 100.10 sdb 0.00 0.00 0.00 11.00 0.00 88.00 0.00 44.00 8.00 0.03 2.27 1.73 1.90 sda 0.00 0.00 220.00 8.00 2792.00 80.00 1396.00 40.00 12.60 146.62 604.96 4.39 100.10 sdb 0.00 0.00 0.00 9.00 0.00 72.00 0.00 36.00 8.00 0.01 1.33 1.11 1.00 sda 1.98 0.00 236.63 0.00 2637.62 0.00 1318.81 0.00 11.15 146.99 571.04 4.19 99.11 sdb 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 sda 0.00 0.00 248.00 0.00 2216.00 0.00 1108.00 0.00 8.94 112.44 582.19 4.04 100.20 sdb 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 sda 2.97 186.14 223.76 10.89 2249.50 1702.97 1124.75 851.49 16.84 50.87 404.61 4.22 99.11 sdb 0.00 182.18 0.00 19.80 0.00 1710.89 0.00 855.45 86.40 1.77 33.80 5.65 11.19 sda 1.02 231.63 114.29 65.31 1869.39 2644.90 934.69 1322.45 25.14 52.77 197.21 5.69 102.14 sdb 0.00 231.63 0.00 81.63 16.33 2628.57 8.16 1314.29 32.40 17.96 137.01 4.75 38.78 sda 1.96 7.84 143.14 60.78 2368.63 172.55 1184.31 86.27 12.46 81.49 287.92 4.81 98.14 sdb 0.00 7.84 5.88 41.18 54.90 180.39 27.45 90.20 5.00 2.83 221.85 8.15 38.33 Why is sda getting all the IO? -- Per Andreas Buer Linpro AS / Ledende på Linux i Norge - 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
