We run the sysbench 0.5 version on our SUT,found the FileIO test case report data about speed is Mb/Gbps,We consider it should be MB/GBps.Is it right?

Bug #1530016 reported by jackhsu
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
sysbench
Invalid
Undecided
Unassigned
sysbench (Ubuntu)
New
Undecided
Unassigned

Bug Description

[root@localhost 600GB]# sysbench --num-threads=16 --test=fileio --file-test-mode=seqwr run
sysbench 0.5: multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 16
Random number generator seed is 0 and will be ignored

Extra file open flags: 0
128 files, 16Mb each
2Gb total file size
Block size 16Kb
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing sequential write (creation) test
Threads started!

Operations performed: 0 reads, 10000 writes, 12800 Other = 22800 Total
Read 0b Written 156.25Mb Total transferred 156.25Mb (14.018Mb/sec)
  897.12 Requests/sec executed
======================================================================
above are our run-status
and one question why seqwrite has other operation 12800 but seqread doesn't?

Revision history for this message
Alexey Kopytov (akopytov) wrote :

Other operations are fsync() calls, which are only executed for write workloads. In your case, you have to fsync() 128 files after every 100 requests out of 10000 total requests.

Changed in sysbench:
status: New → Invalid
Revision history for this message
jackhsu (jackfox3682) wrote :

Exercise me,I also want to ask the result is show xxxx.xx Mbps,why it's "bit" not "Byte"

Revision history for this message
Alexey Kopytov (akopytov) wrote :

Sorry, I missed the first question. It's actually bytes, megabytes, etc. everywhere. It's just that sysbench is using wrong and inconsistent unit names. Fixing it is on my TODO.

Revision history for this message
jackhsu (jackfox3682) wrote : Re: [Bug 1530016] Re: We run the sysbench 0.5 version on our SUT, found the FileIO test case report data about speed is Mb/Gbps, We consider it should be MB/GBps.Is it right?

Thanks for reply

Alexey Kopytov <email address hidden> 於 2016年1月1日 星期五寫道:

> Sorry, I missed the first question. It's actually bytes, megabytes, etc.
> everywhere. It's just that sysbench is using wrong and inconsistent unit
> names. Fixing it is on my TODO.
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1530016
>
> Title:
> We run the sysbench 0.5 version on our SUT,found the FileIO test case
> report data about speed is Mb/Gbps,We consider it should be MB/GBps.Is
> it right?
>
> Status in sysbench:
> Invalid
>
> Bug description:
> [root@localhost 600GB]# sysbench --num-threads=16 --test=fileio
> --file-test-mode=seqwr run
> sysbench 0.5: multi-threaded system evaluation benchmark
>
> Running the test with following options:
> Number of threads: 16
> Random number generator seed is 0 and will be ignored
>
>
> Extra file open flags: 0
> 128 files, 16Mb each
> 2Gb total file size
> Block size 16Kb
> Periodic FSYNC enabled, calling fsync() each 100 requests.
> Calling fsync() at the end of test, Enabled.
> Using synchronous I/O mode
> Doing sequential write (creation) test
> Threads started!
>
> Operations performed: 0 reads, 10000 writes, 12800 Other = 22800 Total
> Read 0b Written 156.25Mb Total transferred 156.25Mb (14.018Mb/sec)
> 897.12 Requests/sec executed
> ======================================================================
> above are our run-status
> and one question why seqwrite has other operation 12800 but seqread
> doesn't?
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/sysbench/+bug/1530016/+subscriptions
>

Revision history for this message
Alexey Kopytov (akopytov) wrote :

Unit have been fixed in 1.0 by this patch: https://github.com/akopytov/sysbench/pull/32

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.