[Moosefs-users] 回复： 回复： Wrong disk usage in mfs cgi monitor and "df -h"

[b. <ai...@qq...>]

Thanks you , Davies .
  
 I've read the xfs wikipedia entry , and it is really impressive , I will do my examinations and post the result  later
   
  
  ------------------ 原始邮件 ------------------
  发件人: "Davies Liu"<dav...@gm...>;
 发送时间: 2011年4月22日(星期五) 凌晨0:53
 收件人: "boblin"<ai...@qq...>; 
 抄送: "youngcow"<you...@gm...>; "Michal Borychowski"<mic...@ge...>; "moosefs-users"<moo...@li...>; 
 主题: Re: [Moosefs-users] 回复： Wrong disk usage in mfs cgi monitor and "df -h"

  
Hi, Bobbin,  

 XFS may be better. We use it as default fs. 

 Davies 
 
  在 2011-4-21，下午8:45， boblin 写道：

  Hi michal & youngcow :
  
 Actually I almost reinstall the whole entire MooseFS system ,and still got the same result.
  
 Now I think may be what youngcow had said is right (mke2fs take 5% space).
  
 At the very fist time (just after I mount the mfs system and before I create any files) ,It alreay cost 5%,
  
 but the whole /data1/chunkdata only cost 2MB disk space .
  
 chunk-140-242:/data1/chunkdata # find /data1/chunkdata/ -type f
/data1/chunkdata/.lock
chunk-140-242:/data1/chunkdata # 
  
 chunk-140-242:/data1/chunkdata # ls -al /data1/chunkdata/* | awk '{ a+=$5; } END {print a/1024/1024"MB";}'
2MB
chunk-140-242:/data1/chunkdata # 
  
  
 <694...@1C...g>
  
 As we know , mke2fs will steal 5% precent space which reservced for superblock . 
  
 Could it be the reason ?
  
 master-140-244:~ # mke2fs -j /dev/cciss/c0d1p1
mke2fs 1.38 (30-Jun-2005)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
89604096 inodes, 179177276 blocks
8958863 blocks (5.00%) reserved for the super user
First data block=0
5469 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks: 
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 
        4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968, 
        102400000
 Writing inode tables: done                            
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done
 This filesystem will be automatically checked every 35 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.
master-140-244:~ # 
   
  
  ------------------ 原始邮件 ------------------
  发件人: "youngcow"<you...@gm...>;
 发送时间: 2011年4月21日(星期四) 下午4:04
 收件人: "Michal Borychowski"<mic...@ge...>; 
 抄送: "'boblin'"<ai...@qq...>; "moosefs-users"<moo...@li...>; 
 主题: Re: [Moosefs-users] Wrong disk usage in mfs cgi monitor and "df -h"

  
I think the reason is filesystem uses some space when you format it.

> Hi!
>
> That really looks a bit strange. We’d blame difference between 'df -h
> /data1' and 'du -sh /data1/chunkdata' on the operating system, and it
> has nothing to do with MooseFS (maybe some difference between protocols
> in kernel or some tools?). What is your operating system? If ‘df’ and
> ‘du’ give wrong values it may be cause of other inconsequences.
>
> And when you make this command: "ls -al /data1/chunkdata/*/*" – how many
> “chunk…mfs” files do you have? What is the sum of their sizes? Do
> something like this:
>
> ls -al /data1/chunkdata/*/* | awk '{ a+=$5; } END {print a;}'
>
> it should give the sum of files sizes in bytes.
>
> Regards
>
> Michal
>
> *From:*boblin [mailto:ai...@qq...]
> *Sent:* Wednesday, April 20, 2011 7:18 PM
> *To:* Michal Borychowski
> *Subject:* Re:RE: RE: [Moosefs-users] Wrong disk usage in mfs cgi
> monitor and "df -h"
>
> hi Michal ：
>
> Now I've got a problem about disk usage , "df -h" shows that partition
> "/data1" used space is 130M . But "du -hs *" shows only
>
> 1.1M。And the most crazy thing is that it show 35GB in cgi monitor ！！
> All my 3 chunkserver have the same problem .
>
> I had read the FAQ and found the "df -h" question ( N * 256 * num_of_hdd
> increased by disk usage），it doesn't help ...
>
> How can I know the exactlly disk usage ?
>
>
>
> ------------------------------------------------------------------------------
> Benefiting from Server Virtualization: Beyond Initial Workload
> Consolidation -- Increasing the use of server virtualization is a top
> priority.Virtualization can reduce costs, simplify management, and improve
> application availability and disaster protection. Learn more about boosting
> the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
>
>
>
> _______________________________________________
> moosefs-users mailing list
> moo...@li...
> https://lists.sourceforge.net/lists/listinfo/moosefs-users




------------------------------------------------------------------------------
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev_______________________________________________
moosefs-users mailing list
moo...@li...
https://lists.sourceforge.net/lists/listinfo/moosefs-users