4902465

During a node board boot, if you have not configured the node board for multiple console use, the message cvc: WARNING: cvc_register: register w/ no console open is displayed.

You can safely ignore this message. If you want to configure the node board for multiple console use, refer to "Configuring Your System for Multiple Console Use" in the Netra CT 820 Server System Administration Guide.

4903566

If you break out of booting the Solaris operating environment and then issue the go command at the ok prompt, you see the message Fast Data Access MMU Miss, and the boot stops.

At the ok prompt, issue the reset command. If the auto-boot? variable is set to false, then issue a boot command after the reset has completed.

4912844

After multiple console use is enabled at the OpenBoot PROM level, the prtpicl command output shows that serial port 0 is unknown, even though a tip session is active.

No workaround available for this release.

4959719

On the Netra CP2300 cPSB board, both the system (boot) flash and the user flash memory reside on the same physical device. If you attempt to write data to the board's user flash PROM while you are reading data from either the user or system flash memory, you may render the board unbootable.

Make sure that you do not access the Netra CP2300 cPSB board's user flash memory for reading and writing simultaneously. You can read and write user flash memory data as separate tasks, but do not attempt to do these read and write operations at the same time.

5016029

If you have connected an external SCSI DAT tape drive to a Netra CP2300 node board, and you perform a reconfiguration boot on the board, the PICL daemon, picld(1M), will terminate and create a core(4) file.

To work around this issue, you must add the following line to the Netra CP2300 board's /etc/system file:

forceload: drv/st

After adding this line, you must reboot the board to apply the change.

Adding this line will force the operating system to load the SCSI disk device driver (sd) kernel module during the kernel initialization. For more information, refer to the system(4) and sd(7D) online man pages.

5027667

If you use the platform information and control library (PICL) software to change the temperature threshold settings of a Netra CP2300 node board, the MOH software will not display the new values of the threshold settings. Instead, the MOH software will display the previous temperature threshold values.

Use the MOH software to set the temperature threshold settings of the Netra CP2300 node boards.

5065036

If you manually stop and start the ctmgx MOH agent running on a node board in rapid succession, the MOH agent running on the distributed management card may not represent the node board correctly.

After stopping the agent, wait for at least 30 seconds before restarting it.

5069151

After connecting external SCSI devices to a Netra CP2300 board and rebooting the board, the MOH software and the prtpicl command displays error messages. Also, the SCSI devices will not be shown in the MOH hierarchy.

To work around this issue, you must add the following line to the Netra CP2300 board's /etc/system file:

forceload: drv/sd

After adding this line, you must reboot the board to apply the change.

Adding this line will force the operating system to load the SCSI disk device driver (sd) kernel module during the kernel initialization. For more information, refer to the system(4) and sd(7D) online man pages.

5070778

In rare situations, after repeated resets or a forced failovers, the distributed management card may panic and display error messages.

The system will recover after the panic without your intervention.

5071230

After connecting external IDE devices to a Netra CP2300 board and rebooting the board, the MOH software and the prtpicl command will display error messages. Also, the IDE devices will not be shown in the MOH hierarchy.

To work around this issue, you must add the following line to the Netra CP2300 board's /etc/system file:

forceload: drv/dad

After adding this line, you must reboot the board to apply the change.

Adding this line will force the operating system to load the IDE disk device driver (dad) kernel module during the kernel initialization. For more information, refer to the system(4) and dad(7D) online man pages.

5073565

When using Netra CT 820 system with the Netra High Availability (HA) Suite Foundation Services Software, the two distributed management cards will not be able to communicate with each other through the carrier grade transport protocol (CGTP) network interfaces.

Before using the Netra CT 820 system with the Netra High Availability (HA) Suite Foundation Services Software, you must replace the /etc/init.d/ctvlan script with the script shown in CODE EXAMPLE 1. See Replacing the ctvlan Initialization Script for more information.

5073598

After using the
cfgadm -c unconfigure command to unconfigure a PCI mezzanine card (PMC) hard drive installed on Netra CP2300 cPSB board, the MOH software will not report the correct status of the PMC drive.

The cfgadm -al command will show that the PMC drive has been unconfigured, but the MOH software will incorrectly show that the drive is still connected and configured.

Use the cfgadm -al command, and not the MOH software, to display the configuration status of the installed devices. Refer to the cfgadm(1M) man page for more information about the command.

5074452

After a distributed management card failover, reset, or system power-on, you cannot set the time of a node board using the ntpdate(1M) command. The ntpdate command will display the following message: no server suitable for synchronization found.

The ntpdate command synchronizes the date and time by polling the network time protocol (NTP) server on the distributed management cards. Once the NTP daemon is started, it can take up to four minutes before the daemon starts providing service to its clients. If a client makes an NTP query within this interval, the client will receive an error message.

5075663

Making an NTP request using the ntpdate(1M) command to a distributed management card alias IP address will not succeed. The ntpdate request returns the message: no server suitable for synchronization found.

When making an NTP request, use the distributed management card's static IP address and not the alias IP address.

To workaround this issue, use the IP addresses of both distributed management cards when using the ntpdate command. For example:

ntpdate 192.168.13.22 192.168.13.23

5076639

When logged into a node board through a distributed management card console session, the node board may reset while using the kadb(1M) kernel debugger.

Distributed management card console sessions are designed for using essential Solaris OS commands, and not for utilities like the kadb kernel debugger. If you wish to use utilities like kadb, log into the node board directly, and not through a console session.