On Sun, Apr 26, 2020 at 7:33 PM Sitsofe Wheeler <sitsofe@xxxxxxxxx> wrote: > > On Sun, 26 Apr 2020 at 14:17, Seena Fallah <seenafallah@xxxxxxxxx> wrote: > > > > Thanks all for your replies. > > > > Maybe if you like it would be a good feature in fio to support this > > type of needs :) > > > > cd /to/filesystem; fio --size=100k --bs=4k --rw=write --name=notabenchmark > > ? Obviously this could go only to the cache so maybe you want > end_fsync=1 (https://fio.readthedocs.io/en/latest/fio_man.html#cmdoption-arg-end-fsync > ) etc. If you really want to do only one I/O I suppose you could > change the block size to 100k... I have tried it but still going to benchmark. I just want to read/write for example 4 IO to see how long does it take? > > I have one more question about xfs_io, I don't know if it's a right > > place but if it not I'm sorry. I have run xfs_io and gets this result: > > 100.000000 bytes, 1 ops; 0.0000 sec (1.514 MiB/sec and 15873.0159 ops/sec) > > What is the 15873.0159 ops/sec? Did xfs_io really do 15873.0159 iops or 1iops? > > What if it did only 1 op but timed how long it took to do all the I/O > (e.g. around 63 microseconds)? When you average that out... The main goal is I am writing a prober for my file system that is attack to a vm for example and I don't want to load on my file system. I just want to probe it and see if it writes 4 IO in a time that was done last time or not? > You can see the code here: > https://git.kernel.org/pub/scm/fs/xfs/xfsprogs-dev.git/tree/io/pwrite.c#n468 I see the code but it wasn't in a pattern I have sent to you. Am I wrong? . > > > On Fri, Apr 24, 2020 at 11:39 AM Yigal Korman <ykorman@xxxxxxxxx> wrote: > > > > > > On Fri, Apr 24, 2020 at 6:16 AM Damien Le Moal <Damien.LeMoal@xxxxxxx> wrote: > > > > > > > > On 2020/04/24 12:11, Damien Le Moal wrote: > > > > > On 2020/04/23 23:41, Seena Fallah wrote: > > > > >> Hi all. > > > > >> > > > > >> I'm trying to probe my file system with fio. I don't want to benchmark > > > > >> my file system. The only thing I want do is to for example write 100K > > > > >> file on a file system and then check how much IO does it take to write > > > > >> and the bandwidth and the runtime. > > > > >> The main thing I want is to just write that file size and don't bench on that. > > > > >> Can anyone help me which ioengine or which args should I use to do this probe? > > > > >> > > > > >> Thanks. > > > > >> > > > > > > > > > > dd if=/dev/zero of=/path/to/your/file/to/write bs=100K count=i conv=fsync > > > > > > > > > > > > > Oops... Should be: > > > > > > > > dd if=/dev/zero of=/path/to/your/file/to/write bs=100K count=1 conv=fsync > > > > > > > > obviously for 100K :) > > > > > > > > You will get the write bandwidth with this. For knowing how many IOs this take, > > > > you will need to trace the kernel IO stack (blktrace). No way to know this > > > > exactly from user space, that is, if by "IO" you mean "storage device commands". > > > > If by "IO" you mean "system calls", then with the above command, it will be > > > > exactly 1 "write()" call (use strace to see it). > > > > > > > > -- > > > > Damien Le Moal > > > > Western Digital Research > > > > > > Yeah, fio might not be the best tool for this purpose. > > > As Damien said - 'dd' is a good alternative. > > > You could also use the xfs_io tool. > > > It's not really xfs related, it's a general i/o and filesystem operations tool. > > > It's part of the xfsprogs package and most likely already installed on > > > your host. > > > Here's an example: > > > > > > xfs_io -c "pwrite 0 100" -f /path/to/file > > > > > > Writes 100 bytes at offset 0 to the file. > > > It will give you some timing info and bandwidth stats. > > > > > > Yigal > > -- > Sitsofe | http://sucs.org/~sits/
