Top
PRIMECLUSTER Global File Services Configuration and Administration Guide 4.5
FUJITSU Software

A.5.3 sfcadm and sfcnode commands

sfcnode: all node entry delete, can't sfcnode

Explanation

The command cannot be executed because information about all nodes sharing the file system will be deleted.

Response

Reduce the specified host information, and re-execute the command.

{sfcadm | sfcnode}: too many host count

Explanation

Too many host names were specified.

Response

Check whether the number of specified host names is correct, and re-execute the command.

{sfcadm | sfcnode}: too many partition count

Explanation

Too many special files were specified.

Response

Check whether the number of specified partitions is correct, and re-execute the command.

{sfcadm | sfcnode}: argument error

Explanation

The specified command format is wrong.

Response

Check the command specification format, and re-execute the command.

{sfcadm | sfcnode}: cannot allocate memory

Explanation

Memory could not be allocated.

Response

Wait for the termination of another process or increase the size of the swap area, and re-execute the command.

{sfcadm | sfcnode}: cannot get host information

Explanation

Acquisition of the host information of node which is specified or which shares file system failed.
Possible causes:

  • The specified node or nodes sharing the file system is stopped.

  • sfcfrmd daemon is not running on the specified node or nodes sharing the file system.

  • A cluster partition error exists.

Response

Execute sfcadm(8) or sfcnode(8) again after correcting the cause of the failure according to the following procedures.

  1. Confirm that the specified node or nodes sharing the file system is not stopped. If the specified node or nodes sharing the file system was stopped, start it.

  2. Execute the following operations on the specified node or nodes sharing the file system.

    1. Check if the sfcfrmd daemon is operating by ps(1) command.
      An example is shown below.

      # /bin/ps -e | /bin/grep sfcfrmd <Enter>
      27198 ?        00:00:00 sfcfrmd
      # echo $? <Enter>
      0

      If grep(1)'s return code is 0, sfcfrmd daemon is running.

      See

      For details of ps(1) and grep(1), see the online manual page.

    2. If the sfcfrmd daemon is not running, confirm that the following message is not output to /var/log/messages.

      WARNING: sfcfsrm:5001: Starting the sfcfrmd daemon was suspended because quorum does not exist

      If the message is not output, execute sfcfrmstart(8) command.

  3. When a problem can not be resolved using operations 1 and 2, a cluster partition error may have occurred. Correct the cluster partition error according to "C.3.4 Corrective action when the sfcfrmd daemon is not started."

{sfcadm | sfcnode}: Cannot connect to FsRM

Explanation

FsRM is not running.

Response

Confirm that FsRM is running.

{sfcadm | sfcnode}: filename cannot open

Explanation

Relevant file (filename) does not exist.

Response

Confirm if filename is available.

Note that, if the management cannot be found, restore it from a backup or create a new one.

{sfcadm | sfcnode}: FSID(fsid_num) is invalid

Explanation

File system identification numbers (fsid_num) are not registered in the management partition.

Response

Check whether the relevant FSID number has been registered with sfcinfo(8).

{sfcadm | sfcnode}: hostid(hostid) not found in volume information

Explanation

The specified host ID (hostid) is not a host ID of the node which shares the GFS Shared File System. Otherwise, the node which executed command is not a node which shares the GFS Shared File System.

Response

Specify the host ID of the node which shares the GFS Shared File System. Otherwise, execute the command again in the node which shares the GFS Shared File System.

{sfcadm | sfcnode}: invalid hostname

Explanation

An invalid host name was specified.

Response

Check the specified host name.

sfcnode: hostname(hostname) is not found

Explanation

The specified host name (hostname) was not found in the node information.

Response

Check whether the specified host name exists in the node information.

{sfcadm | sfcnode}: hostname is not shared

Explanation

The specified node (hostname) does not share the device.
Possible causes:

  • The hostname specified is invalid.

  • The network setting of the hostname is invalid.

  • The hostname is stopped.

  • The sfcfrmd daemon is not running on the hostname.

  • A cluster partition error exists.

Response

Execute sfcadm(8) or sfcnode(8) again after correcting the cause of the failure according to the following procedures.

  1. Check whether the hostname specified is correct. If a mistake is found, specify the correct host name.

  2. On the node specified with the hostname, check whether the IP address of the hostname is correctly registered in host database (such as/etc/hosts). For details, see "11.2.2 Registering information in host database" in this manual.

  3. Confirm that the hostname is not stopped. If the node was stopped, start it.

  4. Execute the following operations on the hostname.

    1. Check if the sfcfrmd daemon is operating by ps(1) command.
      An example is shown below.

      # /bin/ps -e | /bin/grep sfcfrmd <Enter>
      27198 ?        00:00:00 sfcfrmd
      # echo $? <Enter>
      0

      If grep(1)'s return code is 0, sfcfrmd daemon is running.

      See

      For details of ps(1) and grep(1), see the online manual page.

    2. If the sfcfrmd daemon is not running, confirm that the following message is not output to /var/log/messages.

      WARNING: sfcfsrm:5001: Starting the sfcfrmd daemon was suspended because quorum does not exist

      If the message is not output, execute sfcfrmstart(8) command.

  5. When a problem can not be resolved using operations from 1 to 4, a cluster partition error may have occurred. Correct the cluster partition error according to "C.3.4 Corrective action when the sfcfrmd daemon is not started."

sfcnode: hostname registered as MDS, can't sfcnode -d

Explanation

The specified host (hostname) could not delete, because it is registered as the MDS.

Response

Change the MDS host, and re-execute the command.

{sfcadm | sfcnode}: invalid devicename

Explanation

An invalid partition name was specified.

Response

Check whether the specified partition name is correct.

{sfcadm | sfcnode}: invalid magic word magic_no

Explanation

The partition configuration data of the specified partition is invalid.

Response

Check whether the specified partition belongs to a sfcfs file system.

{sfcadm | sfcnode}: ioctl(BLKGETSIZE) error

Explanation

Device information could not be obtained.

Response

Check the system environment to see whether device information is correct.

{sfcadm | sfcnode}: read error (block_no)
{sfcadm | sfcnode}: seek error (
block_no)
{sfcadm | sfcnode}: write error (
block_no)

Explanation

Special file I/O processing failed.

Response

Check whether the specified special file is correct.

{sfcadm | sfcnode}: log partition size too small

Explanation

The specified log size exceeds the size of the entire log partition.

Response

Check the size of the log partition.

{sfcadm | sfcnode}: major number get error

Explanation

The major number could not be obtained.

Response

Check whether the system environment is correct.

sfcadm: invalid MDS hostname

Explanation

The specification of MDS contains an error.

Response

Confirm the management partition and specified MDS information.

{sfcadm | sfcnode}: node full, can't {sfcadm | sfcnode}

Explanation

The number of nodes sharing the file system exceeded the maximum. When the file system is created by sfcmkfs(8), the maximum value is set by the maxnode option.

Response

If you want to change the maximum values of the number of nodes in which the file system is shared, please execute sfcmkfs(8) again, and create the file system again.

{sfcadm | sfcnode}: file system configuration information is full of entry

Explanation

No more file system information could be added.

Response

Delete unnecessary data with sfcadm(8) and re-execute the command.

sfcadm: not enough element of sfcfs partition

Explanation

Not all the partitions belonging to the file system were specified.

Response

Specify all partitions belonging to the file system.

sfcnode: this node entry(hostID:hostid) not found

Explanation

Self-node information was not found in the information for nodes sharing the file system.

Response

Check whether self-node information exists.

{sfcadm | sfcnode}: not super user

Explanation

The command cannot be executed without super-user authority.

Response

Execute the command with super-user authority.

{sfcadm | sfcnode}: not support command

Explanation

The specified command is not supported.

Response

Check the specified command name.

{sfcadm | sfcnode}: now locked, can't {sfcadm | sfcnode}

Explanation

The command could not be executed because another command is being executed.

Response

Upon termination of the other command, re-execute the command.

sfcadm: please specify MDS setting

Explanation

A MDS name must be specified.

Response

Specify a MDS name, and re-execute the command.

sfcadm: please specify PORTNAME setting

Explanation

A service port name must be specified.

Response

Specify a service port name, and re-execute the command.

{sfcadm | sfcnode}: same device is specified more than once

Explanation

A partition that is the same as the one specified was found.

Response

Check whether the specified partition name is correct.

sfcadm: invalid service portname

Explanation

An invalid service port name was specified.

Response

Check the specified service port name.

{sfcadm | sfcnode}: file system configuration information sync error

Explanation

Distribution to another node sharing the file system failed.

Response

Check that a file system watch mechanism works normally, and re-execute the command.

{sfcadm | sfcnode}: special cannot open

Explanation

A special file (special) could not be opened.

Response

Check whether the specified special file is correct.

{sfcadm | sfcnode}: special fstat64 error

Explanation

The status of a special file (special) could not be obtained.

Response

Check whether the specified special file is correct.

{sfcadm | sfcnode}: special is illegal device

Explanation

An invalid partition (special) that is not part of the file system configuration was specified.

Response

Specify all partitions configuring the file system.

{sfcadm | sfcnode}: special is mounted, can't {sfcadm | sfcnode}

Explanation

The command could not be executed because the specified partition (special) has already been mounted.

Response

Unmount the file system, and re-execute the command.

{sfcadm | sfcnode}: special is not block special device

Explanation

The specified file (special) is not a character special file.

Response

Specify a block special file.

{sfcadm | sfcnode}: special is not found in file system configuration information

Explanation

The specified partition (special) is not included in the file system configuration information.

Response

Reexamine whether the specified partition is correct.

{sfcadm | sfcnode}: special is not shared device

Explanation

The specified partition (special) is not shared by nodes.

Response

Check whether the system environment is correct.

{sfcadm | sfcnode}: system error(function_name)

Explanation

A system error occurred.

Response

Contact field engineers.

{sfcadm | sfcnode}: unknown error(error_code)

Explanation

An invalid error was detected.

Response

Contact field engineers.

sfcadm: FSID(fsid_num) is already exist

Explanation

The file system with the file system ID fsid_num has already been created.

Response

Check whether the relevant FSID number has been registered with sfcinfo(8), and re-execute the command.

sfcadm: too many option

Explanation

Overfull options were specified.

Response

Make sure that the number of specified options is proper, and re-execute the command.

{sfcadm | sfcnode}: hostid is illegal hostid

Explanation

The specified host ID is illegal.

Response

Check whether the specified host ID is correct executing sfcsetup(8).

sfcadm: option is illegal option

Explanation

The specified tunable parameter for the file system cannot be registered to the file system configuration information.

Response

Check whether the specified file system's tunable parameter is correct.

{sfcadm | sfcnode}: filename: stat(2) is error: errmsg

Explanation

The stat(2) operation for filename failed.

Response

Take action according to errmsg.

sfcadm: special is already used in file system configuration information

Explanation

The specified partition already exists in the file system configuration information.

Response

Check whether the specified partition is proper.

{sfcadm | sfcnode}: file system configuration information not found

Explanation

The file system configuration information cannot be obtained.

Response

See "Chapter 5 Management Partition" and restore the file system configuration information.

{sfcadm | sfcnode}: special is not sfcfs file system

Explanation

The specified partition (special) is not the GFS Shared File System.

Response

Specify a partition of the GFS Shared File System.

{sfcadm | sfcnode}: ERROR: special is not representative partition

Explanation

The partition special is not a representative partition.

Response

Specify a representative partition then execute the command.

sfcadm: The size of the partition(special) might have been changed after the file system was made

Explanation

The size of the partition (special) which shares the GFS Shared File System might have been changed.

Response

Perform the procedure (from step 2 on) described on "C.3.2 Corrective action in the event of data inconsistency."

GFS Shared File System can not be expanded by size expansion of the partition which shares the GFS Shared File System.

Perform the procedure above promptly.

sfcadm: The size of the partition(special) might have been reduced

Explanation

The size of the partition (special) currently used for a GFS Shared File System may have been reduced.

Response

The file system cannot be repaired. If a backup of the file system is available, recover the file system from the backup.