non7top / ioping Goto Github PK

What steps will reproduce the problem?
1. ioping . -s 4G (actually starting from 2G)


What is the expected output? What do you see instead?

I expect normal workflow, instead I get:
ioping: write failed: Success

What version of the product are you using? On what operating system?

Centos 7.1
3.10.0-123.20.1.el7.x86_64
ioping 0.8

Please provide any additional information below.
At least the output does not lead to anywhere near problem. I have about 20Gig 
free, and it fails to write 2147479552. I use xfs here, and the limit is: 
Maximum File Size = 9 Million TB = 9 ExaB. It may be a problem on my system, 
yet the message given is unclear.

An excerpt from strace:
open("./ioping.jIXgE6", O_RDWR|O_CREAT|O_EXCL, 0600) = 3
unlink("./ioping.jIXgE6")               = 0
pwrite(3, "********************************"..., 4294967296, 0) = 2147479552
write(2, "ioping: ", 8ioping: )                 = 8
write(2, "write failed", 12write failed)            = 12
write(2, ": ", 2: )                       = 2
write(2, "Success\n", 8Success
)                = 8

Alter README to note:

Changes, fixes or improvements to IOPING should be submitted via the 'Issues' 
tab on the ioping project page:

https://code.google.com/p/ioping/issues

What steps will reproduce the problem?
1. Untar 
2. Run make

What is the expected output? What do you see instead?
Program should compile


What version of the product are you using? On what operating system?
ioping-0.4 on Mac OS X 10.6.7

Please provide any additional information below.

make
cc -std=c99 -g -Wall -Wextra -pedantic -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -c 
-o ioping.o ioping.c
ioping.c:26:20: error: malloc.h: No such file or directory
ioping.c: In function ‘parse_device’:
ioping.c:271: warning: implicit declaration of function ‘getline’
ioping.c: In function ‘main’:
ioping.c:338: error: ‘O_DIRECT’ undeclared (first use in this function)
ioping.c:338: error: (Each undeclared identifier is reported only once
ioping.c:338: error: for each function it appears in.)
ioping.c:355: error: ‘BLKGETSIZE64’ undeclared (first use in this function)
ioping.c:375: warning: implicit declaration of function ‘memalign’
ioping.c:375: warning: assignment makes pointer from integer without a cast
ioping.c:413: warning: implicit declaration of function ‘posix_fadvise’
ioping.c:413: error: ‘POSIX_FADV_RANDOM’ undeclared (first use in this 
function)
ioping.c:442: error: ‘POSIX_FADV_DONTNEED’ undeclared (first use in this 
function)
make: *** [ioping.o] Error 1

What steps will reproduce the problem?
1. make
2. ./ioping -R ******
3. doesn't work

What is the expected output? What do you see instead?
The Project Home sample shows that -R is a valid option, but it is not 
available in the actual program.

What version of the product are you using? On what operating system?
0.4 something. Debian Wheezy

Please provide any additional information below.
Please update the documentation to reflect the program's capabilities. It's 
incredibly frustrating to try to use a tool for a documented purpose and then 
have it not actually be able to do that.

Hi,

Can you include a lincese file in the tarball? I'm going to package it into 
Feodra.

Thanks.

Several enhancements for Makefile:

* Drop -g from default CFLAGS, as debugging build shouldn't be the default build
* exchange LDFLAGS with LIBS, LDFLAGS should only contain linker directives, 
where as LIBS contain the libs to be linked with
* Let PREFIX to be easily overwriten

All those changes are necessary for packing ioping easily for distributions.

Please include

What steps will reproduce the problem?
1. run make

What is the expected output? What do you see instead?

Should compile bur returns the following error:

# make
make: Fatal error in reader: Makefile, line 25: Unexpected end of line seen

What version of the product are you using? On what operating system?

ioping-0.8.tar.gz on SunOS 5.11 omnios-33fdde4 i86pc i386 i86p

Please provide any additional information below.

The attached patch is a corrected version of my orginal patch to add OpenBSD 
support. While the original applies and compiles fine, it causes segmentation 
faults. Sorry for the confusion about that one. This corrected version also 
goes a bit further and adds correct labels for partition types and mount point 
names.

What steps will reproduce the problem?
1. compile ioping 0.6
2. run it with -W

What is the expected output? What do you see instead?

Expect it to not throw a usage() error. It does.

Maybe I'm using it wrong? No examples could be found with -W.

What version of the product are you using? On what operating system?

Linux

Please provide any additional information below.

# ioping -WWWL /testfile
ioping: invalid option -- 'W'
 Usage: ioping [-LCDRq] [-c count] [-w deadline] [-p period] [-i interval]
               [-s size] [-S wsize] [-o offset] device|file|directory
        ioping -h | -v

      -c <count>      stop after <count> requests
      -w <deadline>   stop after <deadline>
      -p <period>     print raw statistics for every <period> requests
      -i <interval>   interval between requests (1s)
      -s <size>       request size (4k)
      -S <wsize>      working set size (1m)
      -o <offset>     in file offset
      -L              use sequential operations (includes -s 256k)
      -C              use cached I/O
      -D              use direct I/O
      -R              seek rate test (same as -q -i 0 -w 3 -S 64m)
      -q              suppress human-readable output
      -h              display this message and exit
      -v              display version and exit

# ioping -v
ioping 0.6

I'm experiencing strange behavior when running ioping and using the blkio 
cgroup controller to throttle IOPs. ioping is producing the expected output 
according to what I throttle IOPs at, but it 'hangs' before returning the 
results. I do not see this behavior when not using blkio to throttle IOPs -- 
results are returned much more quickly, as expected.

I am running OpenVZ containers with the latest RHEL6 OpenVZ kernel.

For example, if I run `./ioping -RL .` with IOPs throttled to 150/sec, there is 
a 'pause' for 110 seconds (the length of the pause is pretty consistent), and 
then the results get outputted:

--- . (simfs /vz/private/101) ioping statistics ---
451 requests completed in 3008.0 ms, 153 iops, 38.2 mb/s
min/avg/max/mdev = 1.3/6.5/72.5/15.8 ms

Any thoughts as to why this is happening? I'm trying to determine if this has 
to do with the way ioping is coded, or if it's some sort of anomaly with the 
blkio controller itself.

Compilation of ioping v0.7 is broken under FreeBSD with follow error:

--
ioping.c:579:17: warning: implicit declaration of function 'fdatasync' is 
invalid in C99 [-Wimplicit-function-declaration]
        if (!cached && fdatasync(fd) < 0)
                       ^
2 warnings generated.
/tmp/ioping-QYo1BH.o: In function `do_pwrite':
ioping.c:(.text+0xb16): undefined reference to `fdatasync'
--

cause there no fdatasync functions. fsync(2) must be used instead.


Also, CLANG out the warning about incompatible format for VERSION variable:

--
ioping.c:225:33: warning: format specifies type 'char *' but the argument has 
type 'double' [-Wformat]
        fprintf(stderr, "ioping %s\n", VERSION);
                                ~~     ^~~~~~~
                                %f
<command line>:1:17: note: expanded from here
#define VERSION 0.7
                ^~~
--


Ive don't know what is correct solution to fix this:

-- fprintf(stderr, "ioping %f\n", VERSION);
++ fprintf(stderr, "ioping %f\n", VERSION);

or

-- #define VERSION 0.7
++ #define VERSION "0.7"

or  (char *)VERSION

Temporary ive apply %f type for fprintf.

Please review attached patch.

What steps will reproduce the problem?
1. ioping -c 10 -s 16K /dev/mapper/mpath256
What is the expected output? What do you see instead?

we expected io should happen for 16k on the device (RAW LUN)

Actual Result:

IO's are written on RAW LUNs are 4kb block size.

What version of the product are you using? On what operating system?
[root@DC2-RHEL58-MPIO ~]# ioping -v
ioping 0.6
[root@DC2-RHEL58-MPIO ~]#

OS system: RHEL 5.8


Please provide any additional information below.

Проблема воспроизводилась с 
ioping -v
ioping 0.6
и
./ioping -v
ioping 0.8

Если выполнить ioping -c 10 <mount point>
то время будет фактически одинаковым вне 
зависимости от дисковой системы - 

sudo ./ioping -c 10 -s 1M /
....
10 requests completed in 9.0 s, 233 iops, 233.1 MiB/s

sudo ./ioping -c 10 -s 1M ~/tmpfs/
....
10 requests completed in 9.0 s, 3.2 k iops, 3.1 GiB/s

Показания по iops-ам и скорости разные, но 
время показывает одинаковое и для диска и 
для tmpfs, хотя 10 запросов на tmpfs явно должны 
были пройти быстрее.

-------
We see this problem whith ioping 0.6 and 0.8.
If we run 
ioping -c 10 <mount point>
We always have result in ~9seconds:

sudo ./ioping -c 10 -s 1M /
....
10 requests completed in 9.0 s, 233 iops, 233.1 MiB/s

sudo ./ioping -c 10 -s 1M ~/tmpfs/
....
10 requests completed in 9.0 s, 3.2 k iops, 3.1 GiB/s
But in tmpfs 10 requests must be go for small time, then 9s, if it run 9s on 
hard disk.

Here's a version that seems to be working on Solaris (omnios 5.11). Diff 
attached.

Hello, the output is buffered wich makes it hard to watch and log the stats at 
the same time.
Would it be possible to fflush after each line ? Or maybe add an option in 
front of each line ?
I can provide a patch for each case if needed. Thank you.

I have patched ioping to build and run on OpenBSD. The patch is attached. It 
would be great if the patch could be integrated.

The output of IOPING should use letter-casing that conforms with the usual 
convention of:

B for Bytes, b for bits
M for Mega
etc...

A reported results of "93 mb/s" would display as "93 MBps" or "93 MB/s" etc.

http://en.wikipedia.org/wiki/Data_rate_units#Megabyte_per_second

There is no 0.9 release in Downloads section on the site.

What steps will reproduce the problem?
1. run make

What is the expected output? What do you see instead?

Program should compile, but instead fails with the following compiler message:

# gmake
cc -std=c99 -g -Wall -Wextra -pedantic -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 
-DVERSION=\"0.6\" -c -o ioping.o ioping.c
ioping.c:394:3: error: #error no get disk size method
gmake: *** [ioping.o] Error 1

What version of the product are you using? On what operating system?

ioping-0.6 on DragonFlyBSD 2.13-DEVELOPMENT

Please provide any additional information below.

The issue can be fixed by applying the attached diff.

What steps will reproduce the problem?
1. run make

What is the expected output? What do you see instead?
Program should compile

What version of the product are you using? On what operating system?
ioping-0.4 on FreeBSD 8.1-RELEASE

Please provide any additional information below.
#make
Unknown modifier '/'

Unknown modifier '/'

cc -O2 -pipe  -std=c99 -g -Wall -Wextra -pedantic -D_GNU_SOURCE 
-D_FILE_OFFSET_BITS=64 -c ioping.c
In file included from ioping.c:26:
/usr/include/malloc.h:3:2: error: #error "<malloc.h> has been replaced by 
<stdlib.h>"
ioping.c: In function 'main':
ioping.c:355: error: 'BLKGETSIZE64' undeclared (first use in this function)
ioping.c:355: error: (Each undeclared identifier is reported only once
ioping.c:355: error: for each function it appears in.)
ioping.c:375: warning: implicit declaration of function 'memalign'
ioping.c:375: warning: assignment makes pointer from integer without a cast
ioping.c:413: warning: implicit declaration of function 'posix_fadvise'
ioping.c:413: error: 'POSIX_FADV_RANDOM' undeclared (first use in this function)
ioping.c:442: error: 'POSIX_FADV_DONTNEED' undeclared (first use in this 
function)
*** Error code 1

1. memleak 
2. incorrect strlen