Menu
▾
▴
moosefs-users
|
From: b. <ai...@qq...> - 2011-04-21 17:12:17
|
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 |
|
From: Thomas S H. <tha...@gm...> - 2011-04-21 17:15:02
|
For what it is worth, we also use XFS for all of our chunk servers
2011/4/21 boblin <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
>
>
>
>
> ------------------------------------------------------------------------------
> 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
>
>
|
|
From: Ricardo J. B. <ric...@da...> - 2011-04-21 20:06:25
|
El Jue 21 Abril 2011, Thomas S Hatch escribió: > For what it is worth, we also use XFS for all of our chunk servers Nice to know, today we installed our 10th chunkserver but this time with ext4 (the rest use ext3) and seems good so far :) Will report if > > 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 If reserved space is your concern, for ext3 you reduce it to 50 blocks with: # tune2fs -r 50 /dev/sdb1 And for ext4: # tune4fs -r 50 /dev/sdb1 Of course, replace sdb1 with your real partition name. HTH, -- Ricardo J. Barberis Senior SysAdmin / ITI Dattatec.com :: Soluciones de Web Hosting |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X