Chinaunix

标题: sybase 增加新的设备在新的裸设备，高手请进，帮我看下错误原因啊？ [打印本页]

作者: guojing219 时间: 2009-07-03 10:42
标题: sybase 增加新的设备在新的裸设备，高手请进，帮我看下错误原因啊？
sybase在sco unix  5.0.6下安装成功后，在数据库下面需要增加新的设备在新的裸设备上面。小妹请高手们多多请指点！
我这样建立的啊？怎么报错啊？
我将裸设备划分在第一个区，空间4G应该是够的啊？vdevno=4,4也没有占用啊？到底是什么原因啊？
rsyb_hssb的权限和组我都更改了！列在下面。
我的操作：
1> disk init
2> name='hssb01',
3> physname='/dev/rsyb_hssb',
4> vdevno=4,size=256000
5> go
Msg 5115, Level 16, State 1:
Line 1:
I/O error during disk initialization.  PHYSNAME '/dev/rsyb_hssb' may be
incorrect or 255999 (VSTART + SIZE) may exceed the size of the device.  Please
consult the SQL Server error log for more details.
1>

+-------------------+------------+--------+---+-------------+------------+
| Name             | Type    | New FS | # | First Block | Last Block |
+-------------------+------------+--------+---+-------------+------------+
| syb_hssb       | NON FS    |  no | 0 |          0|    4120199|
|                | NOT USED |  no | 1 |          -|          -|
|                | NOT USED |  no | 2 |          -|          -|
|                | NOT USED |  no | 3 |          -|          -|
|                | NOT USED |  no | 4 |          -|          -|
|                | NOT USED |  no | 5 |          -|          -|
|                | NOT USED |  no | 6 |          -|          -|
| hd1a             | WHOLE DISK |  no | 7 |          0| 143334463|
+-------------------+------------+--------+---+-------------+------------+
143330400 1K blocks for divisions, 4064 1K blocks reserved for the system
n[ame]    Name or rename a division.
c[reate]    Create a new file system on this division.
d[elete]    Delete a file system on this division.
t[ype]    Select or change filesystem type on new filesystems.
p[revent] Prevent a new file system from being created on this division.
s[tart]    Start a division on a different block.
e[nd]       End a division on a different block.
r[estore] Restore the original division table.
# l -l rsyb*
crwxrwxrwx 1 sybase sybase    1,104 Jul  7 09:57 rsyb_hssb

错误日志如下：
00:09/07/07 10:22:11.10 kernel  Initializing virtual device 4, '/dev/rsyb_hssb'
00:09/07/07 10:22:11.10 kernel  Virtual device 4 started using asynchronous i/o.
00:09/07/07 10:22:11.10 kernel  sddone: write error on virtual disk 4 block 0:
00:09/07/07 10:22:11.10 kernel  sddone: No such device or address
00:09/07/07 10:22:11.10 kernel  Deactivating virtual device 4, '/dev/rsyb_hssb'

[qq]65482473[/qq]

[ 本帖最后由 guojing219 于 2009-7-7 10:30 编辑 ]

作者: chuxu 时间: 2009-07-03 11:19
要看报的是什么错误

作者: WFCJZ 时间: 2009-07-03 11:20
能否把报错贴出来?

作者: guojing219 时间: 2009-07-03 11:33
有没有关于在裸设备上建立设备的例子？或者文档呢？数据库装在第一个硬盘，

[ 本帖最后由 guojing219 于 2009-7-7 10:20 编辑 ]

作者: guojing219 时间: 2009-07-03 14:50
如上

[ 本帖最后由 guojing219 于 2009-7-7 10:21 编辑 ]

作者: camham 时间: 2009-07-03 17:37
size＝10240试试，如果不成功再加参数vstart=4096试试

作者: WFCJZ 时间: 2009-07-03 19:55
以下是SYBASE官方网站的，你可以看看，具体情况具体分析！看看有用没！

Disk Errors
This section contains error messages pertaining to disk activation and system administrator disk commands.

--------------------------------------------------------------------------------

Error 5115
Severity
16

Message text
I/O error during disk initialization. PHYSNAME '%.*s' may be incorrect or %ld (VSTART + SIZE) may exceed the size of the device. Please consult the SQL Server error log for more details.

This error may be caused by a hardware problem.

Explanation
This error occurs when a disk init command fails. It can occur for the following reasons:

Incorrect physical device name

Incorrect size

Virtual device number not unique

Permission problem

UNIX kernel not configured for asynchronous I/O.

Action
Check the Adaptive Server error log for information that might help determine the cause of the error.

If necessary, check the items mentioned in the sections below.

Make any needed changes and then run the command again.

Incorrect Device Name
Make sure you have the correct physical device name. Refer to the Adaptive Server installation and configuration guide or the System Administration Guide for instructions.

Incorrect Size
The size parameter of the disk init command must be specified in units of 2K blocks in version 12.0.x and earlier ; or Kilobytes (k or K), Megabytes(m or M), or Gigabytes (g or G) in 12.5 and later.

Sizing 2K units

There are 512 2K blocks in 1MB. Use the formula:

#_of_Megabytes * #_of_data_pages_per_Megabyte = #_of_2K_blocks

For example, to initialize a 10MB Sybase device, specify 5120 for the size:

10MB * 512 = 5120

To determine the number of 2K blocks to use for a raw partition on a UNIX system, convert the size of the partition, in sectors, to 2K blocks. The size of a sector varies on different UNIX systems. For example, it is 512 bytes on Sun, so use sectors/4 for the size parameter. It is 1024 bytes on HP, so use sectors/2. Also, see the information below about virtual device numbers. You must restart Adaptive Server or use a different virtual device number when you reissue the disk init command.

Refer to the Adaptive Server installation and configuration guide for more information about Sybase and your operating system's partitions.

Not all operating systems have partitions as described above.

Virtual Device Number
The vdevno parameter of the disk init command must be unique. Also, if a disk init command fails, Adaptive Server saves the vdevno used, even though the disk init command failed. To make the vdevno available again, restart Adaptive Server. Use sp_helpdevice to determine which virtual device numbers have been used.

The available number of virtual devices is equal to the number of configured devices minus one. For example, if the number of devices configuration parameter is set to 10, then there are 9 available devices, 1 through 9.

Permission Problem
If Adaptive Server was started by the “sybase” account, it must have permission to write to the raw partition or to create a file in the directory specified by disk init. Raw partitions used by Sybase should always be owned by the “sybase” account. Correct the permissions problem and execute the command again, using a different vdevno. Refer to “Virtual Device Number” for more information.

UNIX Kernel
If the UNIX kernel (on SunOS 4.x/Sun Solaris 1.x systems only) is not configured for asynchronous I/O, the 5115 error occurs only if your master device is on a UNIX file, and you are trying to do a disk init on a raw partition. Have your UNIX administrator rebuild the kernel to allow for asynchronous I/O. Refer to Error 823 for more information.

Versions in which this error is raised
All versions

作者: guojing219 时间: 2009-07-07 10:20

原帖由 camham 于 2009-7-3 17:37 发表
size＝10240试试，如果不成功再加参数vstart=4096试试

试过了，还是不行。报的是同样的错误！

作者: hobbylu 时间: 2009-07-07 14:26
你的目标是建一个512M的库，是吧。
给你几条思路
1、确定这个设备上是否有其他的数据库使用这块设备
2、将该设备重建，大小不超过2GB

作者: guojing219 时间: 2009-07-07 16:33
标题: 回复 #9 hobbylu 的帖子
我把裸设备全部删除了，重新建立的，现在还没有建立任何数据库，因为系统是我刚装起来的

作者: hobbylu 时间: 2009-07-07 16:34
将该设备重建，大小不超过2GB

作者: guojing219 时间: 2009-07-08 08:55
标题: 回复 #11 hobbylu 的帖子
你是指裸设备分区吗？

作者: hobbylu 时间: 2009-07-08 09:43
divvy的时候

作者: guojing219 时间: 2009-07-08 10:01

原帖由 hobbylu 于 2009-7-8 09:43 发表
divvy的时候

我改到了2G了，试了下还是不行啊?还是同样的错误.高手能不能留下QQ?帮我看下啊?

作者: guojing219 时间: 2009-07-09 09:55

原帖由 hobbylu 于 2009-7-7 16:34 发表
将该设备重建，大小不超过2GB

在Unixware里无论是采用文件还是裸设备作为数据库的设备，其大小都不能超过2G，因此使用裸设备时，每个分区的大小不能过大，超过2G部分将被浪费

我的系统是sco openserver5.0.6，难道也是这样的吗？
尝试了很多次，最后把裸设备分区改为2G，才能建立设备。

我的问题已经解决了，看样子sco openserver5.0.6和Unixware里一样，都是把裸设备的分区大小分为2G（不超过两G），
我一个5G的数据库只好建立了3个裸设备分区，建立了在3个设备上了。

[ 本帖最后由 guojing219 于 2009-7-9 14:27 编辑 ]

作者: WFCJZ 时间: 2009-07-09 19:55
终于搞好了.恭喜！

欢迎光临 Chinaunix (http://bbs.chinaunix.net/)