Take the 2-minute tour ×
Unix & Linux Stack Exchange is a question and answer site for users of Linux, FreeBSD and other Un*x-like operating systems.. It's 100% free, no registration required.

I'm running tiobench on my raid1 and it gives me the following results.

  1. Why is there ##### shown instead of throughput?
  2. Why does seq.reading with 8 threads use 2836% CPU?

My results:

$tiobench
No size specified, using 2000 MB
Run #1: /usr/bin/tiotest -t 8 -f 250 -r 500 -b 4096 -d . -TTT

Unit information
================
File size = megabytes
Blk Size  = bytes
Rate      = megabytes per second
CPU%      = percentage of CPU used during the test
Latency   = milliseconds
Lat%      = percent of requests that took longer than X seconds
CPU Eff   = Rate divided by CPU% - throughput per cpu load

Sequential Reads
                              File  Blk   Num                   Avg      Maximum      Lat%     Lat%    CPU
Identifier                    Size  Size  Thr   Rate  (CPU%)  Latency    Latency      >2s      >10s    Eff
---------------------------- ------ ----- ---  ------ ------ --------- -----------  -------- -------- -----
2.6.32-37-server              2000  4096    1  ###### 93.62%     0.001        0.02   0.00000  0.00000  6897
2.6.32-37-server              2000  4096    2  ###### 384.4%     0.001        0.03   0.00000  0.00000  2778
2.6.32-37-server              2000  4096    4  ###### 1552.%     0.001        0.07   0.00000  0.00000   980
2.6.32-37-server              2000  4096    8  ###### 2836.%     0.002       10.15   0.00000  0.00000   533

Random Reads
                              File  Blk   Num                   Avg      Maximum      Lat%     Lat%    CPU
Identifier                    Size  Size  Thr   Rate  (CPU%)  Latency    Latency      >2s      >10s    Eff
---------------------------- ------ ----- ---  ------ ------ --------- -----------  -------- -------- -----
2.6.32-37-server              2000  4096    1  ######     0%     0.001        0.01   0.00000  0.00000     0
2.6.32-37-server              2000  4096    2  ######     0%     0.001        0.00   0.00000  0.00000     0
2.6.32-37-server              2000  4096    4  ######     0%     0.001        0.00   0.00000  0.00000     0
2.6.32-37-server              2000  4096    8  ######     0%     0.001        0.00   0.00000  0.00000     0

Sequential Writes
                              File  Blk   Num                   Avg      Maximum      Lat%     Lat%    CPU
Identifier                    Size  Size  Thr   Rate  (CPU%)  Latency    Latency      >2s      >10s    Eff
---------------------------- ------ ----- ---  ------ ------ --------- -----------  -------- -------- -----
2.6.32-37-server              2000  4096    1   87.39 6.816%     0.030     2509.77   0.00059  0.00000  1282
2.6.32-37-server              2000  4096    2   89.32 17.86%     0.057     2478.84   0.00039  0.00000   500
2.6.32-37-server              2000  4096    4   86.98 36.70%     0.113     2353.49   0.00020  0.00000   237
2.6.32-37-server              2000  4096    8   86.46 82.35%     0.214     6847.69   0.00078  0.00000   105

Random Writes
                              File  Blk   Num                   Avg      Maximum      Lat%     Lat%    CPU
Identifier                    Size  Size  Thr   Rate  (CPU%)  Latency    Latency      >2s      >10s    Eff
---------------------------- ------ ----- ---  ------ ------ --------- -----------  -------- -------- -----
2.6.32-37-server              2000  4096    1    1.75 0.335%     0.001        0.02   0.00000  0.00000   521
2.6.32-37-server              2000  4096    2    1.73 0.221%     0.002        0.03   0.00000  0.00000   781
2.6.32-37-server              2000  4096    4    1.71 0.877%     0.002        0.02   0.00000  0.00000   195
2.6.32-37-server              2000  4096    8    1.74 2.671%     0.002        0.03   0.00000  0.00000    65
share|improve this question

1 Answer 1

up vote 4 down vote accepted

This means stuff overflowed and the calculations are meaningless.

If you want to get meaningful figures on the read test, you need to make sure the file size is much larger than your RAM, otherwise all you're testing is essentially your RAM speed and the kernel's caching algorithms.

Try booting with mem=256M at the end of the kernel boot line, and re-run your test with --size 1000 to get ~1G file size under test.
If you don't hear your disks during the test, either they are really silent (good for you!), or the test you're running isn't testing what you think it is.

share|improve this answer
    
Why booting with mem=256M and not just increasing the file-size with --size 4000? –  Fabian Zeindl Jan 4 '12 at 16:59
    
I don't know how much RAM you have, didn't check how tiobench chooses its default - and verified what I said on a machine with 8G, testing with a 32G file didn't appeal to me very much :) –  Mat Jan 4 '12 at 17:01
    
Knoppix (7.2) thought it best not to start with less than 1024m :) The mem= parameter is a good tip! 123 –  Ярослав Рахматуллин Jan 23 at 20:38

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.