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 > > |