Low nilfs Disk utilization

Low nilfs Disk utilization

[Saloustros Georgios]

[-- Attachment #1.1: Type: text/plain, Size: 1131 bytes --]

Hello all.
For my application, which is written in Java, I need a file system, on top
of Linux software raid-0 (md), that guarantees contiguous allocation of
blocks for a large number of small files, in order to increase write
throughput.
For this reason, in the last days I have been experimenting with nilfs2,
which seems the best solution for my application needs.
I am running Red Hat Enterprise Linux 5 (rhel5), kernel-version 2.6.18-prep,
nilfs-2.0.16 and nilfs-utils-2.0.14.
Software Raid-0 consists of two Maxtor 500 GB with chuck size 64K.
System's Memory is 2GB.
For testing reasons I wrote a simple disk benchmark in java, where 8 threads
append 256KB dummy blocks from memory to a log-file. Total IO is 8 GB.
With ext3 mounted with data=writeback mode, each drive alone gives 63MB/sec
and as a raid-0 array a total of 90 MB/sec.

Nilfs has worst performance for the same scenario. 1 Disk alone gives
34MB/sec and two disks on raid-0 setup 39 MB/sec.

Using iostat I can see that disks are 50% utilized with nilfs, whereas with
ext3 are 100% utilized. Why is this happening? Is my configuration wrong?

-- 
Thanks,
G.S.

[-- Attachment #1.2: Type: text/html, Size: 1208 bytes --]

[-- Attachment #2: Type: text/plain, Size: 158 bytes --]

_______________________________________________
users mailing list
users-JrjvKiOkagjYtjvyW6yDsg@public.gmane.org
https://www.nilfs.org/mailman/listinfo/users

Re: Low nilfs Disk utilization

[Jiro SEKIBA]

Hi, 

At Mon, 21 Sep 2009 16:54:50 +0300,
Saloustros Georgios wrote:

> Software Raid-0 consists of two Maxtor 500 GB with chuck size 64K.
> System's Memory is 2GB.
> For testing reasons I wrote a simple disk benchmark in java, where 8 threads
> append 256KB dummy blocks from memory to a log-file. Total IO is 8 GB.
> With ext3 mounted with data=writeback mode, each drive alone gives 63MB/sec
> and as a raid-0 array a total of 90 MB/sec.
> 
> Nilfs has worst performance for the same scenario. 1 Disk alone gives 34MB/sec
> and two disks on raid-0 setup 39 MB/sec.
> 
> Using iostat I can see that disks are 50% utilized with nilfs, whereas with
> ext3 are 100% utilized. Why is this happening? Is my configuration wrong? 

I'm not sure about the utilization but, the performance may be degraded
by the nilfs GC process.  As far as I see with my benchmark,  result is
twice better when stopping the GC process intentionally.

So if that's your case, performance will be expected as  38MB/sec for 1 disk
and 78MB/sec for raid-0.

Or you may tune your GC process by editing /etc/nilfs_cleanerd.conf,
decreasing nsegments_per_clean, increasing cleaning_interval.
Then GC process will affect the results less than it does.

Still, there is more room to improve write performance of nilfs2, I think :).

thanks

regards,
-- 
Jiro SEKIBA <jir-hfpbi5WX9J54Eiagz67IpQ@public.gmane.org>