The FreeBSD Diary

Providing practical examples since 1998

If you buy from Amazon USA, please support us by using this link.

Things look quiet here. But I've been doing a lot of blogging at dan.langille.org because I prefer WordPress now. Not all my posts there are FreeBSD related. I am in the midst of migrating The FreeBSD Diary over to WordPress (and you can read about that here). Once the migration is completed, I'll move the FreeBSD posts into the new FreeBSD Diary website.

rawio - testing device performance 10 August 1999

Need more help on this topic? Click here
This article has no comments
Show me similar articles

After I installed vinum, I wanted to compare the performance of the vinum and non-vinum volumes. I was told about Greg Lehay's benchmark program, rawio.

From man rawio: rawio tests the speed of the low-level character I/O device special in a concurrent environment. It is intended for comparisons of storage devices on a single system, and is not suited for cross-platform performance testing.

rawio is available from ftp://ftp.lemis.com/pub/rawio.tar.gz

Installing rawio

First I fetched the tarball. The -P specifies passive mode as I am behind a firewall. I placed this tarball in /usr/ports/benchmarks/rawio and untar'd it, and installed it

cd /usr/ports/benchmark
mkdir rawio
fetch -P ftp://ftp.lemis.com/pub/rawio.tar.gz
tar xrvf rawio.tar.gzz
make
make install 
WARNING - destructive tests

rawio has the potential to delete data from your disks. Do not run it on volumes which contain precious data unless you have read an understood the man pages first. In particular, the following tests in this article are destructive. Use with caution.

running rawio

rawio has the potential to delete data from your disks. Do not run it on volumes which contain precious data unless you have read an understood the man pages first. In particular, the following tests in this article are destructive. Use with caution.

Here are the tests for the single disks.

Do not run these tests on devices which are mounted. It makes more sense to run these tests on the raw device rather than the

# rawio -a /dev/rda1s1e
        Random read  Sequential read    Random write Sequential write
ID       K/sec  /sec    K/sec  /sec     K/sec  /sec     K/sec  /sec
anon     874.8    54   2683.1   164     868.0    54    1380.3    84

I'm in the process of redoing these tests with a new set of drives.

# rawio -a -v 1 /dev/rda1s1e
Test    ID           K/sec          /sec %User    %Sys  %Total
RR     anon          321.4           20    0.1     8.8     8.9  16384
SR     anon         5518.8          337    0.9    85.8    86.7  16384
RW     anon          403.6           25    0.2    10.7    10.9  16384
SW     anon          617.1           38    0.1    16.4    16.5  16384

# rawio -a -v 1 /dev/da2s1e
Test    ID           K/sec          /sec %User    %Sys  %Total
RR     anon          271.0           17    0.1     7.2     7.3  16384
SR     anon         5290.5          323    0.9    90.0    90.9  16384
RW     anon          514.1           32    0.2    13.1    13.3  16384
SW     anon         1454.7           89    0.2    27.8    28.0  16384

Here is the output from the vinum volume I created.

# rawio -a -v 1 /dev/vinum/test 
Test    ID           K/sec          /sec %User    %Sys  %Total
RR     anon          480.8           30    0.2    18.6    18.9  16384
SR     anon         5001.7          305    0.9    88.8    89.7  16384
RW     anon          478.2           29    0.2    15.4    15.6  16384
SW     anon         1244.2           76    0.2    53.6    53.7  16384

Need more help on this topic? Click here
This article has no comments
Show me similar articles