4TB存储的问题，请大家帮帮，谢谢！！

leonchan0

由于是做文件存储，所以购置了一台磁阵，是4个T的，做的RAID5。

在AS4当中，是用的ext3 。

由于FDISK无法分大于2T的分区，所以当时分区采用的是PARTED  来分的区。

整个4T磁阵都是挂在一个目录下的。

现在的问题是，该目录下有很多个子目录，使用了没两天，就有可能出现部分文件夹内的内容消失了，我用SECURE CRT 远程连接进去，发现这些丢失的文件，在不停的闪烁。有些出现问题的目录下，文件根本就不见了。

但是重启机器后又正常了，说明文件本身并没有丢失，一般在48小时内又出现问题，但是无法读取的内容位置不一定，有可能今天A，C，F目录出现问题，等两天又是 B ，D出问题。都是重启机器后就正常了。

现在我无法判断出错误的位置，有点无从下手，请大家帮忙分析下。

ailms

4TB 的东西lz也敢用 ext3 ，复了

hongling1122

你4T分一个区，太强了……

leonchan0

原帖由 ailms 于 2007-7-4 12:09 发表于 2楼  
4TB 的东西lz也敢用 ext3 ，复了

那用什么呢？

leonchan0

如果用 LVM 来管理 这个问题会不会解决呢？

cnny

ZFS怎么样。不知道LINUX支持不？

leonchan0

ZFS 好象还没有吧

leonchan0

关于阵列的系统日值

Jul  4 08:16:04 localhost kernel: scsi: Device offlined - not ready after error recovery: host 1 channel 0 id 0 lun 0
Jul  4 08:16:04 localhost kernel: scsi: Device offlined - not ready after error recovery: host 1 channel 0 id 0 lun 0
Jul  4 08:16:04 localhost kernel: SCSI error : <1 0 0 0> return code = 0x6000000
Jul  4 08:16:04 localhost kernel: end_request: I/O error, dev sdb, sector 44234
Jul  4 08:16:04 localhost kernel: Buffer I/O error on device sdb1, logical block 5525
Jul  4 08:16:04 localhost kernel: lost page write due to I/O error on sdb1
Jul  4 08:16:04 localhost kernel: Aborting journal on device sdb1.
Jul  4 08:16:04 localhost kernel: SCSI error : <1 0 0 0> return code = 0x6000000
Jul  4 08:16:04 localhost kernel: end_request: I/O error, dev sdb, sector 3991798578
Jul  4 08:16:04 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:16:04 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:16:04 localhost kernel: ext3_abort called.
Jul  4 08:16:04 localhost kernel: EXT3-fs error (device sdb1): ext3_journal_start_sb: Detected aborted journal
Jul  4 08:16:04 localhost kernel: Remounting filesystem read-only
Jul  4 08:16:34 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:16:34 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:16:34 localhost kernel: Buffer I/O error on device sdb1, logical block 498925570
Jul  4 08:16:34 localhost kernel: lost page write due to I/O error on sdb1
Jul  4 08:16:35 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:17:17 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:17:17 localhost kernel: EXT3-fs error (device sdb1): ext3_get_inode_loc: unable to read inode block - inode=249462797, block=498925570
Jul  4 08:54:17 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:56:38 localhost last message repeated 3 times
Jul  4 08:56:38 localhost kernel: EXT3-fs error (device sdb1): ext3_get_inode_loc: unable to read inode block - inode=249462808, block=498925570
Jul  4 08:58:14 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:58:14 localhost kernel: EXT3-fs error (device sdb1): ext3_find_entry: reading directory #244989953 offset 0
Jul  4 08:58:14 localhost kernel:
Jul  4 08:59:26 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 08:59:26 localhost kernel: EXT3-fs error (device sdb1): ext3_find_entry: reading directory #509755393 offset 0
Jul  4 08:59:26 localhost kernel:
Jul  4 09:01:01 localhost crond(pam_unix)[10896]: session opened for user root by (uid=0)
Jul  4 09:01:01 localhost crond(pam_unix)[10896]: session closed for user root
Jul  4 09:45:01 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 09:45:01 localhost kernel: EXT3-fs error (device sdb1): ext3_find_entry: reading directory #321339393 offset 0
Jul  4 09:45:01 localhost kernel:
Jul  4 09:45:14 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 09:45:14 localhost kernel: EXT3-fs error (device sdb1): ext3_get_inode_loc: unable to read inode block - inode=45236261, block=90472451
Jul  4 09:45:25 localhost kernel: scsi1 (0:0): rejecting I/O to offline device
Jul  4 09:45:25 localhost kernel: EXT3-fs error (device sdb1): ext3_find_entry: reading directory #524697601 offset 0
Jul  4 09:45:25 localhost kernel:
Jul  4 09:45:25 localhost kernel: scsi1 (0:0): rejecting I/O to offline device

dhm007

我们以前在freebsd用一个3.6TB的存储空间，也出现读写问题，后面就不接存储了。

建议将4tb分成4个1TB，或8个512g，在通过lvm来管理，有条件可以试试