Troubleshooting Farm Operations

Configuration exception 

Invalid Farm Markup Language (FML) for the farm. 

Contact http://sun.com/service/contacting/index.html for assistance. 

Dynamic Host Configuration Protocol (DHCP) exception 

Cannot create DHCP configuration for the farm. 

This problem can occur if the host name of the control plane server cannot be determined. Make sure the server has the proper network configuration.  

Domain Name Service (DNS) exception 

Cannot create the DNS configuration for the farm. 

This problem can occur if the host name of the control plane server cannot be determined. Make sure the server has the proper network configuration.  

This problem also can occur if the database connectivity is lost. To verify that the control plane database (CPDB) is running and is accepting connections, run any command, such as farm -l. If no proper output displays, the CPDB is not running.

NoMoreResources exception

Not enough resources are available to allocate the farm. 

Check the message to see which resource is exhausted. Use one of the following methods to provision more resources 

• Add resources and add their information to the database

• Free up resources by deactivating existing farms

• Flex down a farm to free up the required devices. Restart the farm operation.

I/O exception 

An I/O error occurred while performing a task such as manipulating disks, configuring monitoring, or accessing files. 

Depending on the specific place where the exception is thrown, take appropriate action that should be part of the message with the exception, such as checking network connections. 

SQL exception 

A database error occurred. 

This problem can occur if the database connectivity is lost. To verify that the CPDB is running and is accepting connections, run any command, such as farm -l. If no proper output displays, the CPDB is not running.

Report all other database errors to http://sun.com/service/contacting/index.html. 

Exceptions other than SQL, such asIllegalState, IllegalArgument, IllegalAccess, and so on

Internal error. 

Contact http://sun.com/service/contacting/index.html for assistance. 

Farm activation fail. 

Check the log file for any critical errors that might point to the cause. 

Depending on what type of error message was received, take appropriate action to activate the farm. 

Farm activation fails during allocation because not enough subnets are available 

The subnet size might be too large. 

You have the option to choose the size of the external subnet for the farm. If you select a size that is not currently in the database, you need to add a subnet by running the subnet command.

Farm deployment fails because named could not restart

The JVM^TM software caches the configuration of the nsswitch.conf file, which describes which database to use for host lookups. If DNS was not part of the nsswitch.conf file's host entry at the time the segment manager was started, all host lookups that cannot be resolved using the /etc/hosts file will fail. See the tspr.debug log for a detailed message describing the error.

1. Ensure that the entry for hosts lookup in the /etc/nsswitch.conf file reads as follows:

   hosts: files dns

2. Restart the segment manager by running the command

   /opt/terraspring/sbin/sm -start

3. Reactivate the farm.

Server is booting, but is unable to get its IP address through DHCP. 

The DHCP daemon might not be running or the media access control (MAC) address of the server is incorrect. 

1. Ensure that the details listed in the database are correct for that server.

   Run the device -lv device-ID command and check the MAC addresses and switch ports.

2. Verify that the DHCP daemon is running and is answering requests by first running ps -ef |grep dhcp, then look in the tspr.debug file to see if there are DHCP messages logged.

3. Connect to the switch and ensure that those ports are connected. Depending on the switch, run either the sh cam dyn port or sh.mac dyn command to ensure that the correct MAC address of the server appears.

4. Check that the Ethernet interface on the control plane server appears as connected on the switch and that it is running as an 802.1q trunk port, with a native virtual local area network (VLAN) of 1.

The control plane server cannot create the DHCP configuration for the farm and receives a message indicating an unknown host. 

The network configuration of the control plane server might be incorrect. 

1. Check the network configuration of the control plane server.

2. Check database connectivity and the file system that contains the DNS configuration.

The control plane server cannot create the DNS configuration for the farm and receives a message indicating an unknown host. 

The network configuration of the control plane server might be incorrect. 

1. Check the network configuration of the control plane server.

2. Check database connectivity and the file system that contains the DNS configuration.

The Control Center does not display a farm correctly after it has been updated. 

Two possible causes: 

1. The update request has not yet completed in the CPDB. This may take a few minutes.

2. The update request has failed.

1. Reconfigure the farm in the Editor dialog and resubmit the update request.

2. Reissue the the farm -af farm ID command from the command line. Then submit another farm update request from the Control Center Editor dialog.

Replace device requests are generated intermittently even though the devices are running and able to ping successfully.

The Ethernet port speeds might not be set to the correct value. 

Ensure that the Ethernet port speeds and duplex setting are set to the same values on all sides (control plane server, switch, and device).