| PRIMECLUSTER Global File Services Configuration and Administration Guide 4.2 (Solaris(TM) Operating Environment) |
|
Contents
Index
|
| Appendix E List of Messages | > E.5 Management Command messages Specific to GFS Shared File System |
Execution was attempted with other than super user permission.
Please execute sfcsetup(1M) with the super user permission.
Communication with the cluster control could not be established.
Check whether the cluster control is operating normally.
Because the lock for exclusion cannot be acquired, it is not possible to change. The sfcfrmd daemon is operating on either of node in the cluster domain. Or, it is thought while sfcsetup(1M) is acquiring the lock on either of node in the cluster domain.
Please stop the sfcfrmd demon with sfcfrmstop(1M) on all nodes of the cluster domain. Or, please wait for the processing completion of sfcsetup(1M). Afterwards, please execute sfcsetup(1M).
Because other commands of the GFS shared file system are executing, it is not possible to change.
Please wait for the completion of other commands of the GFS shared file system, and unmount all file systems if necessary. Afterwards, please stop the sfcfrmd daemon with sfcfrmstop(1M) on all nodes in the cluster domain, and execute sfcsetup(1M).
Too many arguments have been specified.
Please correct the mistake of the argument, and execute sfcsetup(1M).
Too few arguments have been specified. -f option should specify -c option together at the same time.
Check the command format, and then re-execute.
The command format is invalid.
Check the command format, and then re-execute.
The CIP address for the node executing the command could not be acquired.
By remaking and restoring the management partition, back the configuration of the partition to a previous one before a value of the CIP address changed. For details to remake and restore, see Section 20.5, " Backup of the management partition information." and Section 20.6, "Restoration of the management partition information."
The management partition contain invalid data.
Contact local Customer Support.
Failed to write into /var/opt/FJSVsfcfs/sfcfsrm.conf.
Make sure that /var directory has a space by df(1M) command. Make sure that /var/opt/FJSVsfcfs directory exists. If it dose not exist, it is possible to fail to install the FJSVsfcfs package of GFS Shared File System. In this case, install the FJSVsfcfs package again. Otherwise, collect the diagnostic data with fjsnap and contact local Customer Support.
The management partition could not be opened.
Check whether the path to the management partition, specified in the sfcfsrm.conf file, is correct.
The maximum possible amount of node information has already been stored.
Delete information of an unnecessary node by sfcsetup(1M) command with -d option and do it again.
An error occurred when trying to write new node information.
Contact local Customer Support.
The format of an argument specified with the command is invalid.
Check the format of the argument, and then re-execute the command.
An argument specified with the command contains an invalid character string.
Check the argument, and then re-execute the command.
The command cannot be executed, either because another command is being executed, or because sfcfrmd daemon is operating.
Wait for execution of the other command to terminate or, after executing sfcfrmstop(1M) on all the nodes, re-execute the command.
Node (hostid=0xhostid) has still file system information in management partition. It is configured as sharing node of a GFS shared file system. sfcsetup -d did not do anything.
Delete node information by following procedure:
Communication could not be initialized.
Contact local Customer Support. Or, review your system environment.
Communication with the sfcfrmd daemon failed.
Use ps(1) and check whether the sfcfrmd daemon exists. If the daemon does not exist, either execute sfcfrmstart(1M) or execute sfcsetup(1M) in one of the nodes in which the sfcfrmd daemon has been started.
No information could be found for a node that was to be deleted.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
An internal error occurred.
Contact local Customer Support.
Memory could not be allocated. Memory of system is insufficient.
Please check the state of a system, and wait for the termination of another process, or increase the swap space or real memory. After it and re-execute the command.
Reading of sfcfsrm.conf failed.
Collect the diagnostic data with fjsnap and contact local Customer Support.
The file system recovery manager daemon is running on a node for which an attempt was made to execute a command.
Execute sfcfrmstop(1M), and then re-execute the command.
The cluster system may not be operating normally
Check the state and the error message of the cluster system, and operate the cluster system normally.
An input error was made.
None.
The specified character device (raw_device_file) has already been initialized as a management partition.
After checking whether this device really should be initialized, re-execute the command with the -f option specified.
Reading of the management partition failed.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
Writing to the management partition failed.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
Initialization of the management partition failed.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
Either the specified device (raw_device_file) does not exist, or is not a GDS volume.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
The /etc/vfstab file could not be opened.
Make sure that /etc/vfstab exists. If not, restore it from a backup, or see Section 22.2.3, "Setting vfstab" and redefine it.
The device specified (raw_device_file) in the management partition may already be being used by the system.
Either specify a different device in the management partition or, if an entry in /etc/vfstab is not needed, delete it and then re-execute the command.
An invalid row is included in /etc/vfstab.
Confirm the contents of /etc/vfstab and correct it if necessary.
The /etc/mnttab file could not be opened and may be lost.
The problem of Solaris OE occurs when generated in the multiuser mode. Please investigate and deal with the message of Solaris OE.
The device specified (raw_device_file) in the management partition may already be being used by the system.
Either specify a different device in the management partition or, if an entry in the /etc/mnttab file is not needed, delete it and then re-execute the command.
There was an illegal entry in /etc/mnttab.
Please check whether the entry of which file system is a cause, and unmount it at once. Moreover, correct entry of above file system in /etc/vfstab.
The specified device file (raw_device_file) could not be opened.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
The VTOC information for the disk containing the management information could not be acquired.
Show a state of the management partition by sdxinfo(1) command of GDS and make sure that all of the below conditions hold. (For details of the command, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide.")
- The GDS class that includes the management partition must show "shared" in "TYPE" and must show all the nodes that use GFS Shared File System in "SCOPE."
- The volume that is specified as the management partition must have 40MB or more capacity. (In other words, the value of "BLOCKS" is equal to or greater than 8192.)
- The slice that is specified as the management partition must show "ACTIVE" in "STATUS."
If not, see Chapter 16, "Management Partition" and try the setting of the management partition again. If this message is output when all the conditions above hold, collect the diagnostic data with fjsnap and contact local Customer Support.
The major number of disk containing the management partition could not be obtained.
Check whether the system environment is satisfactory, and whether the specified special file is correct.
The diver name for the disk containing the management partition could not be obtained.
Check whether the system environment is satisfactory, and whether the specified special file is correct.
Node information could not be acquired from the management partition.
Collect the diagnostic data with fjsnap and contact local Customer Support.
The information for the node for which the command was executed has already been deleted.
None.
The specified device (raw_device_file) is other than a character special device.
Check the device name, and then re-execute the command.
The sfcfsrm.conf file contains no entry that indicates the path to the management partition. Either node information has not yet been added to the management partition, or it was on a node that has already been deleted.
Check whether node information has not yet been added to the management partition, or whether the node containing the entry has already been deleted.
The specified special file (raw_device_file) does not exist in the system environment.
Check whether the specified special file is correct.
The size of the device specified in the management partition is too small.
Specify a shared disk with a volume size of 40 MB or more.
No entry of management partition in /var/opt/FJSVsfcfs/sfcfsrm.conf. Node information has not been added to the management partition yet or the management partition has been already deleted.
Restore the /var/opt/FJSVsfcfs/sfcfsrm.conf from the backup when it is destroyed.
Add node information to the management partition using sfcsetup -a, if the node is shared node of GFS Shared File System. If the management partition has not been initialized, it must be initialized by sfcsetup -c before adding of node information.
Due to the reason displayed in the message (msg), opening of the FsRM setup file sfcfsrm.conf failed.
Apply the countermeasures described in the message. Please initialize the management partition by sfcsetup(1M) with -c option when msg is No such file or directory.
The volume of the specified shared disk (raw_device_file) is not initialized for the management partition.
Perform sfcsetup(1M) command with -c option to initialize the volume of the shared disk and then re-execute.
The setup file sfcfsrm.conf cannot be found. Node information may not be registered in the node.
Register node information in the management partition by using the sfcsetup(1M) command.
The volume of the shared disk that has been specified in the management partition may not be valid.
Make sure that the slice that is specified as the management partition shows "ACTIVE" in "STATUS" by sdxinfo(1) command. If not, start the slice by sdxvolume(1M) command of GDS. For details, see "PRIMECLUSTER Global Disk Services Configuration and Administration Guide."
The system call syscall failed.
Apply countermeasures described in errmsg.
Node configuration information could not be added to the management partition because the sfcfrmd daemon was executing recovery processing or the operation status of the added node changed.
Re-execute sfcsetup(1M).
Communication with the sfcfrmd daemon failed.
Collect the diagnostic data with fjsnap for all nodes and contact local Customer Support.
Executing the command by clexec(1M) was unsuccessful in the nodename node.
If the stat termination status is 0, check whether the nodename node specified in sfcsetup(1M) is correct. If the node is correct, re-execute sfcsetup(1M).
If the stat termination status is a value other than 0, check the reference manual for clexec(1M), and apply countermeasures in the stat termination status.
Path name setup of the management partition failed in the nodename node.
In all nodes that operate GFS, use ps(1) to check that the sfcfrmd daemon is operating. If the sfcfrmd daemon is not operating in any node, add node configuration information by executing sfcsetup(1M) without using the -r option.
Use cftool(1M) to check that the nodename node is registered in the cluster. In the nodename node, check that the nodename node entry has been set in the /etc/cip.cf file. If a problem is found in either case, set up the cluster for the nodename node.
For details on cluster setup, see the "PRIMECLUSTER Configuration and Administration Guide."
The nodename node is already adding node configuration information to the management partition.
None.
The CIP host name of the nodename node is not defined in /etc/cip.cf.
Define the CIP host name of nodename node in /etc/cip.cf.
For details on /etc/cip.cf, see the "PRIMECLUSTER Cluster Foundation Configuration and Administration Guide."
The sfcfrmd daemon received an invalid data packet while node configuration information was being added to the management partition.
Collect the diagnostic data with fjsnap for all nodes and contact local Customer Support.
The nodename node has not been incorporated into the cluster system.
Set up the cluster in the nodename node, and then start that node.
For details on cluster setup, see the "PRIMECLUSTER Configuration and Administration Guide."
Invalid information was returned as the execution result of clexec(1M).
Collect the diagnostic data with fjsnap for all nodes and contact local Customer Support.
The -r option cannot be used because the sfcfrmd daemon has not been started.
Execute sfcsetup(1M) without the -r option and add the node configuration information.
The acquisition of domain-wide lock(master, config) is in progress.
Wait for a short while.
This node has already been registered in management partition.
None.
|
Contents
Index
|