How to find the MAX IO a physical disk can support

diskio

How to find the MAX I/O a physical disk can support?

My application is doing I/O, and I can find the actual throughput (Blk_wrtn/s) by using linux commands. But how can I find what is max limit I can reach? I want to know if it can be further loaded.

Best Answer

Obiously using Unix tools is the easiest way to do it. You can measure the max operation by creating a test case and use appropriate tools to measure its perfomance. A good resource can be found here: LINUX - Test READ and WRITE speed of Storage

sudo hdparm -tT /dev/sdX

for example as read test.

And to measure write:

dd if=/dev/random of=<some file on the hd> bs=8k count=10000; sync;

# Hit CONTROL-C after 5 seconds to get results
# 65994752 bytes (66 MB) copied, 21.8919 s, 3.0 MB/s


flag

Note As pointed out in the comments the dd command also measures the performance of the file system and even /dev/random. It does measure the write performance of an environment, that heavily depends on the hard disks performance, though.

Related Solutions

How to determine whether the data is written to disk or cached

Read data is (directly) read from the cache only if it is already there. That implies that cached data was previously accessed by a process and kept in cache. There is no system call or any method for a process to know if some piece of data to be read is already in cache or not.

On the other hand, a process can select if it wants written data to be immediately stored on the disk or only after a variable delay which is the general case. This is done by using the O_SYNC flag when opening the file.

There is also the O_DIRECT flag which when supported force all I/Os to bypass the read and write cache and go directly to the disk.

Finally, the hard-disk itself is free to implement its own cache so even after a synchronous write call has returned, there is no guarantee data is already on the disk platters.

Linux – where can I find the disk latency using iostat ( await, svctm, %util )

You can use iostat -x and check for the await column - per device it shows the total time spent waiting plus the actual handling of the request by the disk. The units here are milli-seconds.

$ iostat -yzx 5
Linux 2.6.32-642.13.1.el6.x86_64 (vagrant1)     04/01/2017      _x86_64_        (1 CPU)

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
           0.00    0.00    0.20    0.00    0.00   99.80

Device:         rrqm/s   wrqm/s     r/s     w/s   rsec/s   wsec/s avgrq-sz avgqu-sz   await r_await w_await  svctm  %util
sda               0.00     0.20    0.00    2.20     0.00    19.24     8.73     0.00    2.00    0.00    2.00   0.45   0.10
dm-0              0.00     0.00    0.00    2.40     0.00    19.24     8.00     0.01    2.17    0.00    2.17   0.42   0.10

You can also use sar -d. Again the await column shows avg request latency in ms.

$ sar -d
04:50:01 PM       DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util
05:00:01 PM    dev8-0      0.12      0.00      0.92      7.67      0.00      1.21      0.90      0.01
05:00:01 PM  dev253-0      0.12      0.00      0.92      8.00      0.00      1.45      0.94      0.01
05:00:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
05:10:01 PM    dev8-0      0.14      0.00      1.07      7.90      0.00      1.05      0.80      0.01
05:10:01 PM  dev253-0      0.13      0.00      1.07      8.00      0.00      1.18      0.81      0.01
05:10:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
05:20:01 PM    dev8-0      0.11      0.00      0.79      7.26      0.00      1.52      1.05      0.01
05:20:01 PM  dev253-0      0.10      0.00      0.79      8.00      0.00      2.19      1.15      0.01
05:20:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
05:30:01 PM    dev8-0      0.12      0.00      0.97      7.89      0.00      1.22      0.93      0.01
05:30:01 PM  dev253-0      0.12      0.00      0.97      8.00      0.00      1.42      0.95      0.01
05:30:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
05:40:01 PM    dev8-0      0.12      0.00      0.84      7.20      0.00      0.96      0.77      0.01
05:40:01 PM  dev253-0      0.11      0.00      0.84      8.00      0.00      1.19      0.86      0.01
05:40:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
05:50:01 PM    dev8-0      0.11      0.00      0.84      7.75      0.00      1.31      0.94      0.01
05:50:01 PM  dev253-0      0.11      0.00      0.84      8.00      0.00      2.03      0.97      0.01
05:50:01 PM  dev253-1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

Best Answer

Related Solutions

How to determine whether the data is written to disk or cached

Linux – where can I find the disk latency using iostat ( await, svctm, %util )

Related Question