Disk IO and usage can be checked very easily via ssh, currently the bandwidth usage is 82mbit. IO wait can be checked using the ‘vmstat 1‘ command and looking on the pre-last column called ‘wa’:
procs ———–memory———- —swap– —–io—- –system– —–cpu——
r b swpd free buff cache si so bi bo in cs us sy id wa st
1 1 476 8209760 17060 6923844 0 0 550 22 1 0 10 6 78 7 0
2 2 476 8176176 17152 6934184 0 0 5552 0 7435 2960 26 5 61 8 0
2 0 476 8198232 17160 6938080 0 0 2788 0 6260 2412 14 2 83 1 0
1 1 476 8189900 17184 6947160 0 0 3500 0 7751 2857 22 3 72 3 0
2 0 476 8147572 17192 6963164 0 0 3188 264 10083 3223 24 4 69 3 0
1 0 476 8171132 17208 6966848 0 0 2924 0 5948 2234 18 2 78 2 0
1 1 476 8166708 17220 6972064 0 0 1960 0 6731 2774 15 3 80 2 0
2 1 476 8133964 17236 6982200 0 0 3164 0 8228 3292 18 5 73 4 0
1 0 476 8128256 17252 6986012 0 0 3716 0 5876 2995 21 12 65 3 0
1 2 476 8102740 17336 6990420 0 0 2420 636 6364 3884 22 12 59 8 0
As you can see, its varies from 0-8 which is normal, when IO more than 20-30 for long time means IO is in moderate usage, when its over 50 it is high.
Another command to check disk utilization is ‘iostat –dx 5‘, it calculates 5 seconds disk resources utilization:
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util
sda 18.60 9.80 114.00 8.00 8777.60 142.40 73.11 1.65 13.78 5.44 66.32
sda1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
sda2 18.60 9.80 114.00 8.00 8777.60 142.40 73.11 1.65 13.78 5.44 66.32
dm-0 0.00 0.00 131.80 17.80 8696.00 142.40 59.08 1.95 13.27 4.43 66.34
dm-1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util
sda 27.40 5.00 110.20 7.20 8624.00 97.60 74.29 2.06 17.53 5.88 69.02
sda1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
sda2 27.40 5.00 110.20 7.20 8624.00 97.60 74.29 2.06 17.53 5.88 69.02
dm-0 0.00 0.00 138.40 12.20 8648.00 97.60 58.07 2.96 19.65 4.58 69.00
dm-1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util
sda 21.16 5.79 122.36 6.39 9203.99 97.41 72.25 1.88 14.63 5.23 67.33
sda1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
sda2 21.16 5.79 122.36 6.39 9203.99 97.41 72.25 1.88 14.63 5.23 67.33
dm-0 0.00 0.00 142.71 12.18 9180.04 97.41 59.90 2.36 15.25 4.35 67.31
dm-1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
the last column says about 57-65% utilization of the disks, which is normal at 100mbit (the current bandwidth is 100mbit now).
Another test can be done at rush hours to check the utilization, usually SATA disks in RAID 1 configuration can handle 150-250mbit (depends on number of concurrent connections).
Loading ...