文章存档
相关博文
- ·
- ·
- ·
- ·
- ·
- ·
- ·
- ·
- ·
- ·
分类： oracle
2007-06-21 21:49:50
最近遇到一个问题很是麻烦！
使用oracle 10g, aix, nfs.
问题是这样的：
公司备份出来的数据在别的机器上linux (as3u3)上，通过nfs访问让aix的机器上的oracle可以打开来查看。我们只要想象一下aix的oracle数据文件全部在linux机器上就可以了！
现在遇到oracle 10g访问nfs有点问题：
具体步骤如下:
1>. 在到(oracle 10g)本地机器的控制文件, 将控制文件复制到"/aa"下面
2>. 将本地机器设置为nfs server, nfs client端也为本地机器, nfs server将输出"/aa"
3>. 在本地机器上mount "/aa" 为 "/aa1"使用的文件系统为nfs
4>. 将oracle打开时使用的控制文件指定在nfs文件系统"/aa1"下面.
5>. sql>startup mount pfile=initdb10g.ora;
6>. 这样就出现错误了, 具体错误信息见下面.
aix nfs configure:
##########
testcc-test:test:/test>uname -a
aix testcc-test 3 5 000000004c00
修改下面的配置文件：
在“/etc/hosts”中添加： ip_addr tt
在“/etc/exports”中添加： /aa1 -rw,root=tt,access=tt
使用下面命令来启动nfs服务：
startsrc -g nfs
stopsrc -g nfs
lssrc -g nfs
oracle version:
###########
sql> select * from v$version;
banner
----------------------------------------------------------------
oracle database 10g enterprise edition release 10.2.0.2.0 - 64bi
pl/sql release 10.2.0.2.0 - production
core 10.2.0.2.0 production
tns for ibm/aix risc system/6000: version 10.2.0.2.0 - productio
nlsrtl version 0.2.0.2.0 - production
我已经测试了下面的mount参数，出现的错误信息是一样的。nfs command:
###########
mount -v nfs -o vers=3,proto=udp,rsize=8192,wsize=8192 ip_addr:/aa /aa1
mount -v nfs -o vers=3,hard,rsize=32768,wsize=32768 ip_addr:/aa /aa1
mount -v nfs -o rw,bg,hard,intr,proto=tcp,vers=3,rsize=32768,wsize=32768,timeo=600 ip_addr:/aa /aa1
mount -v nfs -o rw,bg,hard,intr,proto=udp,vers=3,rsize=32768,wsize=32768,timeo=600 ip_addr:/aa /aa1
mount -v nfs -o rw,bg,hard,intr,rsize=32768,wsize=32768,timeo=600,vers=3,proto=tcp,noac,sec=sys ip_addr:/aa /aa1
oracle提示的错误信息，在log文件中。error information:
##########
alter database mount
tue jun 19 15:05:19 2007
ora-00202: control file: '/aa1/db10g_contl_01'
ora-27054: nfs file system where the file is created or resides is not mounted
with correct options
additional information: 6
tue jun 19 15:05:19 2007
ora-205 signalled during: alter database mount...
该机器的man mount信息：
##############
makes a file system available for use.
syntax
mount [ -f ] [ -n node ] [ -o options ] [ -p ] [ -r ] [ -v vfsname ] [ -t type | [ device |
node:directory ] directory | all | -a ] [-v [generic_options] special_mount_points ]
node:directory ] directory | all | -a ] [-v [generic_options] special_mount_points ]
description
the mount command instructs the operating system to make a file system available for use at a
specified location (the mount point). in addition, you can use the mount command to build other
file trees made up of directory and file mounts. the mount command mounts a file system
expressed as a device using the device or node:directory parameter on the directory specified by
the directory parameter. after the mount command has finished, the directory specified becomes
the root directory of the newly mounted file system.
specified location (the mount point). in addition, you can use the mount command to build other
file trees made up of directory and file mounts. the mount command mounts a file system
expressed as a device using the device or node:directory parameter on the directory specified by
the directory parameter. after the mount command has finished, the directory specified becomes
the root directory of the newly mounted file system.
only users with root authority or are members of the system group and have write access to the
mount point can issue file or directory mounts. the file or directory may be a symbolic link.
the mount command uses the real user id, not the effective user id, to determine if the user has
appropriate access. system group members can issue device mounts, provided they have write
access to the mount point and those mounts specified in the /etc/filesystems file. users with
root user authority can issue any mount command.
mount point can issue file or directory mounts. the file or directory may be a symbolic link.
the mount command uses the real user id, not the effective user id, to determine if the user has
appropriate access. system group members can issue device mounts, provided they have write
access to the mount point and those mounts specified in the /etc/filesystems file. users with
root user authority can issue any mount command.
users can mount a device provided they belong to the system group and have appropriate access.
when mounting a device, the mount command uses the device parameter as the name of the block
device and the directory parameter as the directory on which to mount the file system.
when mounting a device, the mount command uses the device parameter as the name of the block
device and the directory parameter as the directory on which to mount the file system.
if you enter the mount command without flags, the command displays the following information for
the mounted file systems:
* the node (if the mount is remote)
* the object mounted
* the mount point
* the virtual-file-system type
* the time mounted
* any mount options
the mounted file systems:
* the node (if the mount is remote)
* the object mounted
* the mount point
* the virtual-file-system type
* the time mounted
* any mount options
if you specify only the directory parameter, the mount command takes it to be the name of the
directory or file on which a file system, directory, or file is usually mounted (as defined in
the /etc/filesystems file). the mount command looks up the associated device, directory, or file
and mounts it. this is the most convenient way of using the mount command, because it does not
require you to remember what is normally mounted on a directory or file. you can also specify
only the device. in this case, the command obtains the mount point from the /etc/filesystems
file.
directory or file on which a file system, directory, or file is usually mounted (as defined in
the /etc/filesystems file). the mount command looks up the associated device, directory, or file
and mounts it. this is the most convenient way of using the mount command, because it does not
require you to remember what is normally mounted on a directory or file. you can also specify
only the device. in this case, the command obtains the mount point from the /etc/filesystems
file.
the /etc/filesystems file should include a stanza for each mountable file system, directory, or
file. this stanza should specify at least the name of the file system and either the device on
which it resides or the directory name. if the stanza includes a mount attribute, the mount
command uses the associated values. it recognizes five values for the mount attributes:
automatic, true, false, removable, and readonly.
file. this stanza should specify at least the name of the file system and either the device on
which it resides or the directory name. if the stanza includes a mount attribute, the mount
command uses the associated values. it recognizes five values for the mount attributes:
automatic, true, false, removable, and readonly.
the mount all command causes all file systems with the mount=true attribute to be mounted in
their normal places. this command is typically used during system initialization, and the
corresponding mounts are referred to as automatic mounts. note: if the cdromd cd and dvd
their normal places. this command is typically used during system initialization, and the
corresponding mounts are referred to as automatic mounts. note: if the cdromd cd and dvd
automount daemon is enabled, then those devices will be automatically mounted as specified in
the /etc/cdromd.conf file. use the cdumount or cdeject command to unmount an automounted cd or
dvd. use stopsrc -s cdromd to disable the cd/dvd automount daemon. note: for cachefs, the remote
filesystem that is to be cached locally must be exported such that the root id of the local
system is not remapped on the remote host to nobody (or the id that the remote host uses as the
anonymous user). for example, if host a were to export a filesystem /f, which would be mounted
with cachefs on host b, then the /etc/exports on host a would need to have an entry similar to:
the /etc/cdromd.conf file. use the cdumount or cdeject command to unmount an automounted cd or
dvd. use stopsrc -s cdromd to disable the cd/dvd automount daemon. note: for cachefs, the remote
filesystem that is to be cached locally must be exported such that the root id of the local
system is not remapped on the remote host to nobody (or the id that the remote host uses as the
anonymous user). for example, if host a were to export a filesystem /f, which would be mounted
with cachefs on host b, then the /etc/exports on host a would need to have an entry similar to:
/f -rw,root=b
or
/f -ro,root=b
depending on the mount options used for the local cachefs mount. note: mounting a jfs filesystem
on a read-only logical volume is not supported.
or
/f -ro,root=b
depending on the mount options used for the local cachefs mount. note: mounting a jfs filesystem
on a read-only logical volume is not supported.
flags
-a
mounts all file systems in the /etc/filesystems file with stanzas that contain the true
mount attribute.
all
same as the -a flag.
-f
requests a forced mount during system initialization to enable mounting over the root file
system.
-n node
specifies the remote node that holds the directory to be mounted. for nfs version 4 mounts
only, the node can be specified as a colon-separated ipv6 address. if this is done with the
node:directory format, the colon-separated ipv6 address must be enclosed in square
brackets.
mounts all file systems in the /etc/filesystems file with stanzas that contain the true
mount attribute.
all
same as the -a flag.
-f
requests a forced mount during system initialization to enable mounting over the root file
system.
-n node
specifies the remote node that holds the directory to be mounted. for nfs version 4 mounts
only, the node can be specified as a colon-separated ipv6 address. if this is done with the
node:directory format, the colon-separated ipv6 address must be enclosed in square
brackets.
file system specific options
-o options
specifies options. options entered on the command line should be separated only by a comma.
the following file system-specific options do not apply to all virtual file system types:
bsy
prevents the mount operation if the directory to be mounted over is the current
working directory of a process.
cio
specifies the file system to be mounted for concurrent readers and writers. i/o on
files in this filesystem will behave as if they had been opened with o_cio specified
in the open() system call. using this option will prevent access in any manner other
than cio. it is impossible to use cached i/o on a filesystem mounted with the cio
option. this means that mapping commands such as mmap() and shmat() will fail with
einval when used on any file in a filesystem mounted with the cio option. one side-
effect of this is that it is impossible to run binaries out of a cio mounted
filesystem, since the loader may use mmap().
dio
specifies that i/o on the filesystem will behave as if all the files had been opened
with o_direct specified in the open() system call. note: using the -odio or -ocio
flags can help performance on certain workloads, but users should be aware that
using these flags will prevent file caching for these file systems. because
readahead is disabled for these file systems, this may decrease performance for
large sequential reads.
fmode=octal
specifies the mode for a file and directory. the default is 755.
gid=gid
specifies the gid that is assigned to files in the mount. the default is bin.
log=lvname
-o options
specifies options. options entered on the command line should be separated only by a comma.
the following file system-specific options do not apply to all virtual file system types:
bsy
prevents the mount operation if the directory to be mounted over is the current
working directory of a process.
cio
specifies the file system to be mounted for concurrent readers and writers. i/o on
files in this filesystem will behave as if they had been opened with o_cio specified
in the open() system call. using this option will prevent access in any manner other
than cio. it is impossible to use cached i/o on a filesystem mounted with the cio
option. this means that mapping commands such as mmap() and shmat() will fail with
einval when used on any file in a filesystem mounted with the cio option. one side-
effect of this is that it is impossible to run binaries out of a cio mounted
filesystem, since the loader may use mmap().
dio
specifies that i/o on the filesystem will behave as if all the files had been opened
with o_direct specified in the open() system call. note: using the -odio or -ocio
flags can help performance on certain workloads, but users should be aware that
using these flags will prevent file caching for these file systems. because
readahead is disabled for these file systems, this may decrease performance for
large sequential reads.
fmode=octal
specifies the mode for a file and directory. the default is 755.
gid=gid
specifies the gid that is assigned to files in the mount. the default is bin.
log=lvname
specifies the full path name of the filesystem logging logical volume name where the
following file-system operations are logged.
maxpout=value
specifies the pageout level for files on this filesystem at which threads should be
slept. if maxpout is specified, minpout must also be specified. value must be non-
negative and greater than minpout. the default is the kernel maxpout level.
minpout=value
specifies the pageout level for files on this filesystem at which threads should be
readied. if minpout is specified, maxpout must also be specified. value must be
non-negative. the default is the kernel minpout level.
nocase
turns-off case mapping. this is useful for cdroms using the iso 9660:1998/hsg
standard.
nodev
specifies that you cannot open devices from this mount. this option returns a value
of enxio if a failure occurs.
nosuid
specifies that execution of setuid and setgid programs by way of this mount is not
allowed. this option returns a value of eperm if a failure occurs.
rbr
mount filesystem with the release-behind-when-reading capability. when sequential
reading of a file in this filesystem is detected, the real memory pages used by the
file will be released once the pages are copied to internal buffers. note: when rbr
is specified, the d_rb_read flag is ultimately set in the _devflags field in the
pdtentry structure.
rbw
mount filesystem with the release-behind-when-writing capability. when sequential
writing of a file in this filesystem is detected, the real memory pages used by the
file will be released once the pages written to disk. note: when rbw is specified,
the d_rb_write flag is set.
rbrw
mount filesystem with both release-behind-when-reading and release-behind-when-
writing capabilities. note: if rbrw is specified, both the d_rb_read and the
d_rb_write flags are set.
ro
specifies that the mounted file is read-only. the default value is rw.
rw
specifies that the mounted file is read/write accessible. rw is the default value.
snapshot
specifies the device to be mounted is a snapshot. the snapped file system for the
specified snapshot must already be mounted or an error message will display.
snapto=snapshot
specifies the location to start a snapshot with the value of snapshot when mounting
the specified jfs2 file system.
upcase
changes case mapping from default lowercase to uppercase. this is useful for cdroms
using the iso 9660:1998/hsg standard.
uid=uid
specifies the uid that is assigned to files in the mount, the defeult is bin.
wrkgrp=workgroup
specifies the workgroup that the smb server belongs.
nfs specific options
-o options
specifies options. options you enter on the command line should be separated only by a
comma, not a comma and a space. the following nfs-specific options do not apply to all
virtual file system types:
acdirmax=n
holds cached attributes for no more than n seconds after directory update. the
default is 60 seconds.
following file-system operations are logged.
maxpout=value
specifies the pageout level for files on this filesystem at which threads should be
slept. if maxpout is specified, minpout must also be specified. value must be non-
negative and greater than minpout. the default is the kernel maxpout level.
minpout=value
specifies the pageout level for files on this filesystem at which threads should be
readied. if minpout is specified, maxpout must also be specified. value must be
non-negative. the default is the kernel minpout level.
nocase
turns-off case mapping. this is useful for cdroms using the iso 9660:1998/hsg
standard.
nodev
specifies that you cannot open devices from this mount. this option returns a value
of enxio if a failure occurs.
nosuid
specifies that execution of setuid and setgid programs by way of this mount is not
allowed. this option returns a value of eperm if a failure occurs.
rbr
mount filesystem with the release-behind-when-reading capability. when sequential
reading of a file in this filesystem is detected, the real memory pages used by the
file will be released once the pages are copied to internal buffers. note: when rbr
is specified, the d_rb_read flag is ultimately set in the _devflags field in the
pdtentry structure.
rbw
mount filesystem with the release-behind-when-writing capability. when sequential
writing of a file in this filesystem is detected, the real memory pages used by the
file will be released once the pages written to disk. note: when rbw is specified,
the d_rb_write flag is set.
rbrw
mount filesystem with both release-behind-when-reading and release-behind-when-
writing capabilities. note: if rbrw is specified, both the d_rb_read and the
d_rb_write flags are set.
ro
specifies that the mounted file is read-only. the default value is rw.
rw
specifies that the mounted file is read/write accessible. rw is the default value.
snapshot
specifies the device to be mounted is a snapshot. the snapped file system for the
specified snapshot must already be mounted or an error message will display.
snapto=snapshot
specifies the location to start a snapshot with the value of snapshot when mounting
the specified jfs2 file system.
upcase
changes case mapping from default lowercase to uppercase. this is useful for cdroms
using the iso 9660:1998/hsg standard.
uid=uid
specifies the uid that is assigned to files in the mount, the defeult is bin.
wrkgrp=workgroup
specifies the workgroup that the smb server belongs.
nfs specific options
-o options
specifies options. options you enter on the command line should be separated only by a
comma, not a comma and a space. the following nfs-specific options do not apply to all
virtual file system types:
acdirmax=n
holds cached attributes for no more than n seconds after directory update. the
default is 60 seconds.
acdirmin=n
holds cached attributes for at least n seconds after directory update. the default
is 30 seconds.
acl
requests using the access control list rpc program for this nfs mount. if the acl
option is used, the acl rpc program is used only if the nfs server provides it. the
default is noacl.
acregmax=n
holds cached attributes for no longer that n seconds after file modification. the
default is 60 seconds.
acregmin=n
holds cached attributes for at least n seconds after file modification. the default
is 3 seconds.
actimeo=n
sets minimum and maximum times for regular files and directories to n seconds. if
this option is set, it overrides any settings for the acregmin, acregmax, acdirmin,
and acdirmax options.
bg
attempts mount in background if first attempt is unsuccessful. the default value is
fg.
biods=n
sets the maximum number of biod threads that perform asynchronous i/o rpc requests
for an nfs mount. the maximum value that can be set is 128. values greater than 128
are limited to 128 within the nfs client. the nfs client dynamically manages the
number of running biod threads up to the maximum based on activity. the default
maximums for the different nfs protocols are 7 for nfs version 2, 4 for nfs version
3, and 16 for nfs version 4. these defaults are subject to change in future
releases.
cio
specifies the file system to be mounted for concurrent readers and writers. i/o on
files in this filesystem will behave as if they had been opened with o_cio specified
in the open() system call. using this option will prevent access in any manner other
than cio. it is impossible to use cached i/o on a filesystem mounted with the cio
option. this means that mapping commands such as mmap() and shmat() will fail with
einval when used on any file in a filesystem mounted with the cio option. one side-
effect of this is that it is impossible to run binaries out of a cio mounted
filesystem, since the loader may use mmap().
dio
specifies that i/o on the filesystem will behave as if all the files had been opened
with o_direct specified in the open() system call. note: using the -odio or -ocio
flags can help performance on certain workloads, but users should be aware that
using these flags will prevent file caching for these file systems. because
readahead is disabled for these file systems, this may decrease performance for
large sequential reads.
fastattr
bypasses the requirement that files currently being written will be sent to the
server before the attributes of the file is read. this option is to be used with
caution, since it will cause the client to assume that the file data that has not
yet reached the server will be written without problem. in case of write errors, the
client and server will have different opinions on what the size of the file really
is. likewise, a client will not be aware of attribute changes to the file being made
by another client, so this option must not be used in environments where two clients
are writing to the same files.
fg
attempts mount in foreground if first attempt is unsuccessful. fg is the default
value.
grpid
directs any file or directory created on the file system to inherit the group id of
the parent directory.
holds cached attributes for at least n seconds after directory update. the default
is 30 seconds.
acl
requests using the access control list rpc program for this nfs mount. if the acl
option is used, the acl rpc program is used only if the nfs server provides it. the
default is noacl.
acregmax=n
holds cached attributes for no longer that n seconds after file modification. the
default is 60 seconds.
acregmin=n
holds cached attributes for at least n seconds after file modification. the default
is 3 seconds.
actimeo=n
sets minimum and maximum times for regular files and directories to n seconds. if
this option is set, it overrides any settings for the acregmin, acregmax, acdirmin,
and acdirmax options.
bg
attempts mount in background if first attempt is unsuccessful. the default value is
fg.
biods=n
sets the maximum number of biod threads that perform asynchronous i/o rpc requests
for an nfs mount. the maximum value that can be set is 128. values greater than 128
are limited to 128 within the nfs client. the nfs client dynamically manages the
number of running biod threads up to the maximum based on activity. the default
maximums for the different nfs protocols are 7 for nfs version 2, 4 for nfs version
3, and 16 for nfs version 4. these defaults are subject to change in future
releases.
cio
specifies the file system to be mounted for concurrent readers and writers. i/o on
files in this filesystem will behave as if they had been opened with o_cio specified
in the open() system call. using this option will prevent access in any manner other
than cio. it is impossible to use cached i/o on a filesystem mounted with the cio
option. this means that mapping commands such as mmap() and shmat() will fail with
einval when used on any file in a filesystem mounted with the cio option. one side-
effect of this is that it is impossible to run binaries out of a cio mounted
filesystem, since the loader may use mmap().
dio
specifies that i/o on the filesystem will behave as if all the files had been opened
with o_direct specified in the open() system call. note: using the -odio or -ocio
flags can help performance on certain workloads, but users should be aware that
using these flags will prevent file caching for these file systems. because
readahead is disabled for these file systems, this may decrease performance for
large sequential reads.
fastattr
bypasses the requirement that files currently being written will be sent to the
server before the attributes of the file is read. this option is to be used with
caution, since it will cause the client to assume that the file data that has not
yet reached the server will be written without problem. in case of write errors, the
client and server will have different opinions on what the size of the file really
is. likewise, a client will not be aware of attribute changes to the file being made
by another client, so this option must not be used in environments where two clients
are writing to the same files.
fg
attempts mount in foreground if first attempt is unsuccessful. fg is the default
value.
grpid
directs any file or directory created on the file system to inherit the group id of
the parent directory.
hard
retries a request until server responds. the option is the default value.
intr
allows keyboard interrupts on hard mounts.
llock
requests that files lock locally at the nfs client. nfs network file locking
requests are not sent to the nfs server if the llock option is used.
maxgroups=n
indicates that nfs rpc calls using auth_unix may include up to n member groups of
information. using this option to increase the number of member groups beyond the
rpc protocol standard of 16 will only work against servers that support more than 16
member groups. otherwise, the client will experience errors.
retries a request until server responds. the option is the default value.
intr
allows keyboard interrupts on hard mounts.
llock
requests that files lock locally at the nfs client. nfs network file locking
requests are not sent to the nfs server if the llock option is used.
maxgroups=n
indicates that nfs rpc calls using auth_unix may include up to n member groups of
information. using this option to increase the number of member groups beyond the
rpc protocol standard of 16 will only work against servers that support more than 16
member groups. otherwise, the client will experience errors.
values below 16 or greater than 64 will be ignored. by default, the protocol
standard maximum of 16 is adhered to. aix nfs servers will accept and process
auth_unix credentials with up to 64 groups starting with aix 5l version 5.2 with the
5200-01 recommended maintenance package. the actual number of member groups sent by
the nfs client is dependent on the number of groups the involved user is a member
of, and may be limited by the length of the nfs client's hostname (which is included
in the auth_unix information).
noac
specifies that the mount command performs no attribute or directory caching. if you
do not specify this option, the attributes (including permissions, size, and
timestamps) for files and directories are cached to reduce the need to perform
over-the-wire nfsproc_getattr remote procedure calls (rpcs). the nfsproc_getattr rpc
enables a client to prompt the server for file and directory attributes. the
acregmin, acregmax, acdirmin, and acdirmax options control the length of time for
which the cached values are retained.
noacl
specifies not to use the access control list rpc program for this nfs mount request.
the default is noacl.
nointr
specifies no keyboard interrupts allowed on hard mounts.
port=n
sets server internet protocol (ip) port number to n. the default value is the 2049.
posix
requests that pathconf information be exchanged and made available on an nfs version
2 mount. requires a mount version 2 rpc.mountd at the nfs server.
proto=[udp|tcp]
specifies the transport protocol. the default is tcp. use the proto=[udp|tcp] option
to override the default.
standard maximum of 16 is adhered to. aix nfs servers will accept and process
auth_unix credentials with up to 64 groups starting with aix 5l version 5.2 with the
5200-01 recommended maintenance package. the actual number of member groups sent by
the nfs client is dependent on the number of groups the involved user is a member
of, and may be limited by the length of the nfs client's hostname (which is included
in the auth_unix information).
noac
specifies that the mount command performs no attribute or directory caching. if you
do not specify this option, the attributes (including permissions, size, and
timestamps) for files and directories are cached to reduce the need to perform
over-the-wire nfsproc_getattr remote procedure calls (rpcs). the nfsproc_getattr rpc
enables a client to prompt the server for file and directory attributes. the
acregmin, acregmax, acdirmin, and acdirmax options control the length of time for
which the cached values are retained.
noacl
specifies not to use the access control list rpc program for this nfs mount request.
the default is noacl.
nointr
specifies no keyboard interrupts allowed on hard mounts.
port=n
sets server internet protocol (ip) port number to n. the default value is the 2049.
posix
requests that pathconf information be exchanged and made available on an nfs version
2 mount. requires a mount version 2 rpc.mountd at the nfs server.
proto=[udp|tcp]
specifies the transport protocol. the default is tcp. use the proto=[udp|tcp] option
to override the default.
proto=udp cannot be specified if vers=4.
retrans=n
sets the number of nfs transmissions to n. the default value is 5. the retrans
setting determines how many times the nfs client retransmits a given udp rpc request
to an nfs server for file system operations. the retrans setting is not used during
communication with the nfs server rpc.mountd service when processing nfs version 2
and 3 mounts. retries to rpc.mountd are controlled with the retry mount option.
retry=n
sets the number of times the mount is attempted to n; the default value is 1000.
when the retry value is 0, the system makes 10,000 attempts.
rsize=n
sets the read buffer size to n bytes. the default value is 8192. for aix 4.2.1 and
later, the default value is 32768 when using version 3 of the nfs protocol. for aix
5.3 and later, the default value is 32768 when using version 4 of the nfs protocol.
secure
specifies that the mount command uses data encryption standard (des) for nfs
transactions.
retrans=n
sets the number of nfs transmissions to n. the default value is 5. the retrans
setting determines how many times the nfs client retransmits a given udp rpc request
to an nfs server for file system operations. the retrans setting is not used during
communication with the nfs server rpc.mountd service when processing nfs version 2
and 3 mounts. retries to rpc.mountd are controlled with the retry mount option.
retry=n
sets the number of times the mount is attempted to n; the default value is 1000.
when the retry value is 0, the system makes 10,000 attempts.
rsize=n
sets the read buffer size to n bytes. the default value is 8192. for aix 4.2.1 and
later, the default value is 32768 when using version 3 of the nfs protocol. for aix
5.3 and later, the default value is 32768 when using version 4 of the nfs protocol.
secure
specifies that the mount command uses data encryption standard (des) for nfs
transactions.
sec=flavor[:flavor...]
specifies a list of security methods that may be used to access files under the
mount point. allowable flavor values are:
sys
unix(r) authentication. this is the default method.
dh
des authentication.
krb5
kerberos. authentication only.
krb5i
kerberos. authentication and integrity.
krb5p
kerberos. authentication, integrity, and privacy.
specifies a list of security methods that may be used to access files under the
mount point. allowable flavor values are:
sys
unix(r) authentication. this is the default method.
dh
des authentication.
krb5
kerberos. authentication only.
krb5i
kerberos. authentication and integrity.
krb5p
kerberos. authentication, integrity, and privacy.
the secure option may be specified, but not in conjunction with a sec option.
the secure option is deprecated and may be eliminated in a future release.
use sec=dh instead.
sec=[flavor1:...:flavorn]
the sec option specifies the security flavor list for the nfs mount. the available
flavors are des, unix, sys, krb5, krb5i, and krb5p. this option only applies to aix
5.3 or later.
shortdev
specifies that you are mounting a file system from a host that does not support 32-
bit device special files.
soft
returns an error if the server does not respond. the default value is hard.
timeo=n
sets the network file system (nfs) time-out period to n tenths of a second. for tcp
mounts, the default timeout is 100, which equals 10 seconds. for udp mounts, the
default timeout is 11, which equals 1.1 seconds, but varies depending on the nfs
operation taking place. for udp mounts, the timeout will increase for each failed
transmission, with a maximum value of 20 seconds. each transmission will be
attempted twice, after which the timeout value is updated. the timeo option does not
apply to communication from the nfs client to the rpc.mountd service on nfs servers.
a timeout of 30 seconds is used when making calls to rpc.mountd.
vers=[2|3|4]
specifies nfs version. the default is the version of nfs protocol used between the
client and server and is the highest one available on both systems. if the nfs
server does not support nfs version 3, the nfs mount will use nfs version 2. use the
vers=[2|3|4] option to select the nfs version. by default, the nfs mount will never
use nfs version 4 unless specified. the vers=4 only applies to aix 5.3 or later.
wsize=n
sets the write buffer size to n bytes. the default value is 8192. for aix 4.2.1 and
later, the default value is 32768 when using version 3 of the nfs protocol. for aix
5.3 and later, the default value is 32768 when using version 4 of the nfs protocol.
-p
mounts a file system as a removable file system. while open files are on it, a removable
mounted file system behaves the same as a normally mounted file system. however, when no
files are open (and no process has a current directory on the file system), all of the file
system disk buffers in the file system are written to the medium, and the operating system
forgets the structure of the file system.
-r
mounts a file system as a read-only file system, regardless of its previous specification
in the /etc/filesystems file.
-t type
mounts all stanzas in the /etc/filesystems file that contain the type=type attribute and
are not mounted. the type parameter specifies the name of the group.
-v vfsname
the secure option is deprecated and may be eliminated in a future release.
use sec=dh instead.
sec=[flavor1:...:flavorn]
the sec option specifies the security flavor list for the nfs mount. the available
flavors are des, unix, sys, krb5, krb5i, and krb5p. this option only applies to aix
5.3 or later.
shortdev
specifies that you are mounting a file system from a host that does not support 32-
bit device special files.
soft
returns an error if the server does not respond. the default value is hard.
timeo=n
sets the network file system (nfs) time-out period to n tenths of a second. for tcp
mounts, the default timeout is 100, which equals 10 seconds. for udp mounts, the
default timeout is 11, which equals 1.1 seconds, but varies depending on the nfs
operation taking place. for udp mounts, the timeout will increase for each failed
transmission, with a maximum value of 20 seconds. each transmission will be
attempted twice, after which the timeout value is updated. the timeo option does not
apply to communication from the nfs client to the rpc.mountd service on nfs servers.
a timeout of 30 seconds is used when making calls to rpc.mountd.
vers=[2|3|4]
specifies nfs version. the default is the version of nfs protocol used between the
client and server and is the highest one available on both systems. if the nfs
server does not support nfs version 3, the nfs mount will use nfs version 2. use the
vers=[2|3|4] option to select the nfs version. by default, the nfs mount will never
use nfs version 4 unless specified. the vers=4 only applies to aix 5.3 or later.
wsize=n
sets the write buffer size to n bytes. the default value is 8192. for aix 4.2.1 and
later, the default value is 32768 when using version 3 of the nfs protocol. for aix
5.3 and later, the default value is 32768 when using version 4 of the nfs protocol.
-p
mounts a file system as a removable file system. while open files are on it, a removable
mounted file system behaves the same as a normally mounted file system. however, when no
files are open (and no process has a current directory on the file system), all of the file
system disk buffers in the file system are written to the medium, and the operating system
forgets the structure of the file system.
-r
mounts a file system as a read-only file system, regardless of its previous specification
in the /etc/filesystems file.
-t type
mounts all stanzas in the /etc/filesystems file that contain the type=type attribute and
are not mounted. the type parameter specifies the name of the group.
-v vfsname
specifies that the file system is defined by the vfsname parameter in the /etc/vfs file.
cachefs specific options
the cachefs-specific version of the mount command mounts a cached file system; if necessary, it
nfs-mounts its back file system. it also provides a number of cachefs-specific options for
controlling the caching process.
nfs-mounts its back file system. it also provides a number of cachefs-specific options for
controlling the caching process.
to mount a cachefs file system, use the mount command with the -v flag followed by the argument.
the following mount flags are available.
the following mount flags are available.
the following arguments to the -o flag are specifically for cachefs mounts. options you enter on
the command line should be separated only by a comma, not a comma and a space. note: the
backfstype argument must be specified.
-o
specifies options.
acdirmax=n
specifies that cached attributes are held for no more than n seconds after directory
update. before n seconds, cachefs checks to see if the directory modification time
on the back file system has changed. if it has, all information about the directory
is purged from the cache and new data is retrieved from the back file system. the
default value is 60 seconds.
acdirmin=n
specifies that cached attributes are held for at least n seconds after directory
update. after n seconds, cachefs checks to see if the directory modification time on
the back file system has changed. if it has, all information about the directory is
purged from the cache and new data is retrieved from the back file system. the
default value is 30 seconds.
acregmax=n
specifies that cached attributes are held for no more than n seconds after file
modification. after n seconds, all file information is purged from the cache. the
default value is 30 seconds.
acregmin=n
specifies that cached attributes are held for at least n seconds after file
modification. after n seconds, cachefs checks to see if the file modification time
on the back file system has changed. if it has, all information about the file is
purged from the cache and new data is retrieved from the back file system. the
default value is 30 seconds.
actimeo=n
sets acregmin, acregmax, acdirmin, and acdirmax to n.
backfstype=file_system_type
the file system type of the back file system (for example, nfs).
backpath=path
specifies where the back file system is already mounted. if this argument is not
supplied, cachefs determines a mount point for the back file system.
cachedir=directory
the name of the cache directory.
cacheid=id
id is a string specifying a particular instance of a cache. if you do not specify a
cache id, cachefs will construct one.
demandconst
enables maximum cache consistency checking. by default, periodic consistency
checking is enabled. when you enable demandconst, it checks on every read and write.
note: note: if this option is used the first time a specific cachefs is mounted,
then the option must also be specified for subsequent mounts. there is state
information stored in the cache control files that enforces consistent use of this
option.
local_access
the command line should be separated only by a comma, not a comma and a space. note: the
backfstype argument must be specified.
-o
specifies options.
acdirmax=n
specifies that cached attributes are held for no more than n seconds after directory
update. before n seconds, cachefs checks to see if the directory modification time
on the back file system has changed. if it has, all information about the directory
is purged from the cache and new data is retrieved from the back file system. the
default value is 60 seconds.
acdirmin=n
specifies that cached attributes are held for at least n seconds after directory
update. after n seconds, cachefs checks to see if the directory modification time on
the back file system has changed. if it has, all information about the directory is
purged from the cache and new data is retrieved from the back file system. the
default value is 30 seconds.
acregmax=n
specifies that cached attributes are held for no more than n seconds after file
modification. after n seconds, all file information is purged from the cache. the
default value is 30 seconds.
acregmin=n
specifies that cached attributes are held for at least n seconds after file
modification. after n seconds, cachefs checks to see if the file modification time
on the back file system has changed. if it has, all information about the file is
purged from the cache and new data is retrieved from the back file system. the
default value is 30 seconds.
actimeo=n
sets acregmin, acregmax, acdirmin, and acdirmax to n.
backfstype=file_system_type
the file system type of the back file system (for example, nfs).
backpath=path
specifies where the back file system is already mounted. if this argument is not
supplied, cachefs determines a mount point for the back file system.
cachedir=directory
the name of the cache directory.
cacheid=id
id is a string specifying a particular instance of a cache. if you do not specify a
cache id, cachefs will construct one.
demandconst
enables maximum cache consistency checking. by default, periodic consistency
checking is enabled. when you enable demandconst, it checks on every read and write.
note: note: if this option is used the first time a specific cachefs is mounted,
then the option must also be specified for subsequent mounts. there is state
information stored in the cache control files that enforces consistent use of this
option.
local_access
causes the front file system to interpret the mode bits used for access checking
instead or having the back file system verify access permissions. do not use this
argument with secure nfs.
noconst
disables cache consistency checking. by default, periodic consistency checking is
enabled. specify noconst only when you know that the back file system will not be
modified. trying to perform cache consistency check using cfsadmin-s will result in
error. demandconst and noconst are mutually exclusive. note: note: if this option is
used the first time a specific cachefs is mounted, then the option must also be
specified for subsequent mounts. there is state information stored in the cache
control files that enforces consistent use of this option.
purge
purge any cached information for the specified file system. note: note: if this
option is used the first time a specific cachefs is mounted, then the option must
also be specified for subsequent mounts. there is state information stored in the
cache control files that enforces consistent use of this option.
rw | ro
read-write (default) or read-only.
suid | nosuid
allow (default) or disallow set-uid execution
write-around | non-shared
write modes for cachefs. the write-around mode (the default) handles writes the same
as nfs does; that is, writes are made to the back file system, and the affected file
is purged from the cache. you can use the non-shared mode when you are sure that no
one else will be writing to the cached file system. note: note: if this option is
used the first time a specific cachefs is mounted, then the option must also be
specified for subsequent mounts. there is state information stored in the cache
control files that enforces consistent use of this option.
-v
mounts a cachefs file system.
instead or having the back file system verify access permissions. do not use this
argument with secure nfs.
noconst
disables cache consistency checking. by default, periodic consistency checking is
enabled. specify noconst only when you know that the back file system will not be
modified. trying to perform cache consistency check using cfsadmin-s will result in
error. demandconst and noconst are mutually exclusive. note: note: if this option is
used the first time a specific cachefs is mounted, then the option must also be
specified for subsequent mounts. there is state information stored in the cache
control files that enforces consistent use of this option.
purge
purge any cached information for the specified file system. note: note: if this
option is used the first time a specific cachefs is mounted, then the option must
also be specified for subsequent mounts. there is state information stored in the
cache control files that enforces consistent use of this option.
rw | ro
read-write (default) or read-only.
suid | nosuid
allow (default) or disallow set-uid execution
write-around | non-shared
write modes for cachefs. the write-around mode (the default) handles writes the same
as nfs does; that is, writes are made to the back file system, and the affected file
is purged from the cache. you can use the non-shared mode when you are sure that no
one else will be writing to the cached file system. note: note: if this option is
used the first time a specific cachefs is mounted, then the option must also be
specified for subsequent mounts. there is state information stored in the cache
control files that enforces consistent use of this option.
-v
mounts a cachefs file system.
examples
1 to list the mounted file systems, enter:
1 to list the mounted file systems, enter:
mount
this command produces output similar to the following:
node mounted mounted vfs date options over
---- ------- --------- --- ------------ ------- ---------
/dev/hd0 / jfs dec 17 08:04 rw, log =/dev/hd8
/dev/hd3 /tmp jfs dec 17 08:04 rw, log =/dev/hd8
/dev/hd1 /home jfs dec 17 08:06 rw, log =/dev/hd8
/dev/hd2 /usr jfs dec 17 08:06 rw, log =/dev/hd8
sue /home/local/src /usr/code nfs dec 17 08:06 ro, log =/dev/hd8
---- ------- --------- --- ------------ ------- ---------
/dev/hd0 / jfs dec 17 08:04 rw, log =/dev/hd8
/dev/hd3 /tmp jfs dec 17 08:04 rw, log =/dev/hd8
/dev/hd1 /home jfs dec 17 08:06 rw, log =/dev/hd8
/dev/hd2 /usr jfs dec 17 08:06 rw, log =/dev/hd8
sue /home/local/src /usr/code nfs dec 17 08:06 ro, log =/dev/hd8
for each file system, the mount command lists the node name, the device name, the name
under which it is mounted, the virtual-file-system type, the date and time it was mounted,
and its options.
2 to mount all default file systems, enter: mount all
under which it is mounted, the virtual-file-system type, the date and time it was mounted,
and its options.
2 to mount all default file systems, enter: mount all
this command sequence mounts all standard file systems in the /etc/filesystems file marked
by the mount=true attribute.
3 to mount a remote directory, enter: mount -n nodea /home/tom.remote /home/tom.local
by the mount=true attribute.
3 to mount a remote directory, enter: mount -n nodea /home/tom.remote /home/tom.local
this command sequence mounts the /home/tom.remote directory located on nodea onto the local
/home/tom.local directory. it assumes the default vfsname parameter=remote, which must be
defined in the /etc/vfs file.
/home/tom.local directory. it assumes the default vfsname parameter=remote, which must be
defined in the /etc/vfs file.
4 to mount a file or directory from the /etc/filesystems file with a specific type, enter:
mount -t remote
mount -t remote
this command sequence mounts all files or directories in the /etc/filesystems file that
have a stanza that contains the type=remote attribute.
5 to cachefs-mount the file system which is already nfs-mounted on /usr/abc, enter:
have a stanza that contains the type=remote attribute.
5 to cachefs-mount the file system which is already nfs-mounted on /usr/abc, enter:
mount -v cachefs -o backfstype=nfs,backpath=/usr/abc,
cachedir=/cache1 server1:/user2 /xyz
cachedir=/cache1 server1:/user2 /xyz
the lines similar to the following appear in the /etc/mnttab file after the mount command
is executed:
is executed:
server1:/user2 /usr/abc nfs
/usr/abc /cache1/xyz cachefs backfstype=nfs
6 to mount a snapshot, type:
/usr/abc /cache1/xyz cachefs backfstype=nfs
6 to mount a snapshot, type:
mount -o snapshot /dev/snapsb /home/janet/snapsb
this command mounts the snapshot contained on the /dev/snapsb device onto the
/home/janet/snapsb directory.
7 to mount a file system and create a snapshot, type:
this command mounts the snapshot contained on the /dev/snapsb device onto the
/home/janet/snapsb directory.
7 to mount a file system and create a snapshot, type:
mount -o snapto=/dev/snapsb /dev/sb /home/janet/sb
this command mounts the file system contained on the /dev/sbdevice onto the /home/janet/sb
directory and creates a snapshot for the file system on the /dev/snapsbdevice.
8 to access files on an smb server as a local filesystem, type:
this command mounts the file system contained on the /dev/sbdevice onto the /home/janet/sb
directory and creates a snapshot for the file system on the /dev/snapsbdevice.
8 to access files on an smb server as a local filesystem, type:
mount -v cifs -n pezman/user1/pass1 -o uid=201,fmode=750 /home /mnt
files
/etc/filesystems
lists the known file systems and defines their characteristics.
/etc/vfs
contains descriptions of virtual-file-system types.
lists the known file systems and defines their characteristics.
/etc/vfs
contains descriptions of virtual-file-system types.
related information
the cdcheck command, cdeject command, cdmount command, cdromd command, cdumount command, cdutil
command, nfso command, umount command.
the mntctl subroutine, mount subroutine, umount subroutine.
the filesystems file, vfs file.
for information on installing the web-based system manager, see chapter 2: installing web-based
system manager in aix 5l version 5.3 web-based system manager administration guide.
mounting and system management interface tool in operating system and device management.
performance implications of hard or soft nfs mounts in performance management.
===================================
直接i/o支持。
以前，oracle数据库仅对oracle集群文件系统（ocfs）提供直接i/o支持。现在，oracle 10g还对网络文件系统（nfs）中的直接i/o提供支持。"有了oracle 10g，就有了直接支持，而且是直接在数据库中提供的，而不是通过打补丁获得的，"coekaerts说。"这可以确确实实地提高性能--尤其是在nfs 中，所有事务处理的速度都变快了。在nfs上运行rac真好。"coekaerts认为，这种直接i/o支持对于单个节点也非常重要，因为你不必将数据缓存在os文件系统中，但却可以使用oracle的算法。
来源：
===================================
该问题解决了[20060626]
参考文档: oracle® real application clusters 10g release 2 with aix 5l version 5.3 and netapp storage
note:
the “cio” mount option should not be used for oracle_home and crs_home (the volume where crs and oracle binaries are placed).
oracle install will fail if you dynamically mount the nfs volumes without adding entries in the /etc/filesystems file.
-o rw,bg,hard,intr,proto=tcp,vers=3,rsize=32768,wsize=32768,timeo=600
$ uname -a
aix p270 2 5 0005a9dc4c00
在文件"/etc/filesystems"中添加的项目有:
/aa1:
dev = /aa
vfs = nfs
log = /tmp/nfslog
mount = true
options = rw
account = false
command, nfso command, umount command.
the mntctl subroutine, mount subroutine, umount subroutine.
the filesystems file, vfs file.
for information on installing the web-based system manager, see chapter 2: installing web-based
system manager in aix 5l version 5.3 web-based system manager administration guide.
mounting and system management interface tool in operating system and device management.
performance implications of hard or soft nfs mounts in performance management.
===================================
直接i/o支持。
以前，oracle数据库仅对oracle集群文件系统（ocfs）提供直接i/o支持。现在，oracle 10g还对网络文件系统（nfs）中的直接i/o提供支持。"有了oracle 10g，就有了直接支持，而且是直接在数据库中提供的，而不是通过打补丁获得的，"coekaerts说。"这可以确确实实地提高性能--尤其是在nfs 中，所有事务处理的速度都变快了。在nfs上运行rac真好。"coekaerts认为，这种直接i/o支持对于单个节点也非常重要，因为你不必将数据缓存在os文件系统中，但却可以使用oracle的算法。
来源：
===================================
该问题解决了[20060626]
参考文档: oracle® real application clusters 10g release 2 with aix 5l version 5.3 and netapp storage
note:
the “cio” mount option should not be used for oracle_home and crs_home (the volume where crs and oracle binaries are placed).
oracle install will fail if you dynamically mount the nfs volumes without adding entries in the /etc/filesystems file.
-o rw,bg,hard,intr,proto=tcp,vers=3,rsize=32768,wsize=32768,timeo=600
$ uname -a
aix p270 2 5 0005a9dc4c00
在文件"/etc/filesystems"中添加的项目有:
/aa1:
dev = /aa
vfs = nfs
log = /tmp/nfslog
mount = true
options = rw
account = false
给主人留下些什么吧！~~