| PRIMECLUSTER Global File Services Configuration and Administration Guide 4.1 (Solaris(TM) 10 Operating System) |
|
Contents
Index
|
| Appendix A List of Messages | > A.3 The GFS Shared File System Script Messages |
file_name is not executable.
See "Installation Guide " of PRIMECLUSTER and reinstall the package.
The status of the GDS volume that was specified as the management partition is not ACTIVE. Or, there is a possibility that the GDS volume is deleted without deleting node configuration information from the management partition by sfcsetup -d on the target node when the use of GFS is stopped.
Check whether the volume exists by executing the following procedures:
# ls -l management_partition <Enter> |
Create a volume according to the instructions in "Operation" of the "PRIMECLUSTER Global Disk Services Configuration and Administration Guide (Solaris(TM) Operating System)".
Restore the management partition according to the instructions in "9.7 Restoring of the management partition information".
See "9.2 Creating of the management partition" and set up the management partition.
If the volume status is not ACTIVE, change the status to ACTIVE. For instructions on how to change the volume status, see the "Operation" of the "PRIMECLUSTER Global Disk Services Configuration and Administration Guide (Solaris(TM) Operating System)".
Delete information on the deleted GDS volume by executing the following procedures:
1. Create a temporary GDS volume of 40 megabytes or more. (GDS volume name is arbitrary)
2. sfcsetup -c temporary_GDS_volume (GDS volume name made at 1.)
3. sfcsetup -a temporary_GDS_volume (GDS volume name made at 1.)
4. sfcsetup -d
Delete the node configuration information from the management partition by executing the following procedures:
1. Set the volume status to ACTIVE. For instructions on how to change the volume status, see "Operations" of the "PRIMECLUSTER Global Disk Services Configuration and Administration Guide (Solaris(TM) Operating System)".
2. See "9.4 Deleting node configuration information from the management partition" and delete the node configuration information from the management partition.
sfcfrmd daemon failed booting.
Check the error message of sfcfrmstart(1M) displayed previously to this message, and take corrective action about that message according to "A.5.3 sfcfrmstart command".
The daemon could not be stopped.
Collect a file output with fjsnap for troubleshooting information then contact your customer support engineers.
File access to cancel the suspension of the sfcfrmd daemon startup failed.
Check that the /var/opt/FJSVgfs directory exists. If it does, restart the system. If it doesn't, install the package according to the installation guide.
Checking a quorum exists failed.
Check the message of rcqquery(1M) displayed previously to this message, and after taking corrective action about that message according to "rcqquery messages" of the "PRIMECLUSTER Cluster Foundation (CF) (Solaris) Configuration and Administration Guide" , restart the system.
Activation of the GFS service is disrupted because a quorum does not exist when CF is started from GUI. If the sfcfrmd daemon's start mode is set to wait, and a quorum does not exist, the GFS service will be discontinued during CF startup.
A quorum does not exist, and the GFS service is discontinued under the following conditions:
If you activate CF from GUI after stopping it on all the nodes, see the procedure described in "12.6 How to start up CF from GUI when a GFS Shared File System is used".
If you did not stop CF on all the nodes, or the GFS service activation fails after performing the procedure detailed above, a cluster partition error might have occurred. Take corrective action according to "C.3.3 Corrective action when the sfcfrmd daemon is not started".
The zonename(1) command failed.
Execute the zonename(1) command then check the correct result is output and reboot the system. If the same problem occurs again, collect a file output with fjsnap for troubleshooting information then contact your customer support engineers.
Activation of the sfcfrmd daemon is suspended because a quorum does not exist when the node is started up, the node enters multi-user mode, or CF is started from GUI. Normally, a quorum be not existing is not maintained until all the nodes are activated. As soon as it is exist, the sfcfrmd daemon will be activated.
Accordingly, it is necessary to take corrective actions under the following conditions otherwise data remains inconsistent.
Take corrective action according to "C.3.3 Corrective action when the sfcfrmd daemon is not started".
Since acquisition of the startup mode of the sfcfrmd daemon fails, the daemon is activated with the wait setting.
If GFS is activated properly, no corrective action needs to be taken. If an error message of the sfcfsrm script is output, it is necessary to take corrective action according to the message.
Since a quorum exists, the suspended sfcfrmd daemon was started.
None.
Since sfcfrmstart -f was executed, the suspended sfcfrmd daemon was started.
None.
|
Contents
Index
|