OCFS文件系统出现INODE满的状况,但实际空间利用率没有
查了METALINK,搜索出了这篇,但不知道如何解决啊
Bug 号 3333232
已归档 21-DEC-2003 已更新 29-SEP-2006
产品 Linux OS 产品版本1.0.9
平台 Red Hat Advanced Server 平台版本 无数据
数据库版本 9.2.0.4 影响平台Port-Specific
优先级 Severe Loss of Service 状态 Closed, Not a Bug
基本 Bug N/A 修复产品版本 无数据
问题陈述:
THE RESULT OF 'DF -I' TO OCFS DOES NOT RETRUN EXPECTED VALUE
--------------------------------------------------------------------------------
*** 12/21/03 11:13 pm *** . PROBLEM:The result of 'df -i' to OCFS does not retrun expected value. .
ex. 1. Copying files onto OCFS, IUsed(the number of inodes) decrease.
2. Removing files from OCFS, IUsed incerase. .
Please refer to testcase for details. .
DIAGNOSTIC ANALYSIS:Such a problem does not occur on nomarl device. .
WORKAROUND:n/a .
RELATED BUGS:n/a .
REPRODUCIBILITY:.100% (both In-house and Ct' site)
Rep? Platform
OCFS Ver. -------------------------------------- Y(100%)
RedHat Advanced Server 2.1 OCFS 1.0.9-9Y(100%)
RedHat Advanced Server 2.1 OCFS 1.0.9-12 .
TESTCASE:Confirm the Inode information during coppying and removingfiles with the following shell scripts. .
### conf.sh ### while((1)) do df -i sleep 5 done .
1. copy files onto OCFS . Filesystem
@ OCFS does not have the concept of inodes, however, for the vfs layer in the
@ kernel we need to fill out structures. the number returned by df -i is
@ based on diskspace free. it's the number of clusterblocsk in the filesystem
@ we dont' preallocate inodes like most filesystems do. nor do we HAVE to do
@ that. not a bug, expected behaviour, not required to do what any other
@ filesystem does.
CLosed as NOT a bug.