IBM Books

Installation and Migration Guide


BOS node upgrade

This method is used primarily when you need to:

This method applies AIX PTF Service and preserves the current rootvg disk configuration. |It installs and upgrades the AIX BOS file sets. The pssp_script installs and updates the |PSSP 3.4 LPs on top of current PSSP LPs.

Notes:

  1. If you use this method, only the required PSSP file sets will be upgraded. Optional file sets must be upgraded manually.

  2. For the following tasks, you can elect not to use dsh and issue commands on the node.

Step 1: Apply AIX 4.3.3 upgrade on the node

Issue the NFS mount command from the control workstation using dsh to mount the lppsource directory on the control workstation to the node.

dsh -w node "/usr/sbin/mount \
CWS:/spdata/sys1/install/lppsource_name/lppsource /mnt"

Issue the lslpp and installp commands to update all listed LPs on the node with AIX service found in the lppsource directory. You can issue the commands directly from the node.

dsh -w node "/usr/bin/lslpp -l -J  > /tmp/FILE.43"
dsh -w node "/usr/sbin/installp -acNgXd /mnt -f /tmp/FILE.43"

Step 2: Verify AIX migration

Verify that the node has successfully been migrated to the target AIX level by issuing the following command:

dsh -w node "/bin/oslevel"

For example, if the target AIX level is 4.3.3 and the output of this command does not indicate AIX 4.3.3, issue the following command to return a list of AIX files not migrated to AIX 4.3.3. You may need to install AIX PTFs to migrate those file sets to AIX 4.3.3.

dsh -w node "/bin/oslevel -l 4.3.3.0"

Step 3: Reboot the node

If you just upgraded your node (an AIX modification level change, for example, AIX 4.3.2 to 4.3.3), you now need to reboot the node so changes to the kernel will take effect. To do this, issue:

|cshutdown -rGFN node_list

Step 4: Enter node configuration data

You need to set the appropriate SDR node object attributes for lppsource_name, code_version, bootp_response, and pv_list for each node being upgraded. Use the spchvgobj and spbootins commands to update these fields. If you are migrating nodes in more than one system partition, you will need to issue these commands in each system partition. For a complete description of the flags associated with these commands, refer to PSSP: Command and Technical Reference.

For example, to migrate nodes 1 and 2 to AIX 4.3.3 and |PSSP 3.4 where lppsource was placed in /spdata/sys1/install/aix433/lppsource, issue the following two commands:

|spchvgobj -r selected_vg -p PSSP-3.4 -v aix433 -h 00-00-00-0,0 \
|          -l 1,2 -i bos.obj.ssp.433
| 
|spbootins -s no -r customize -l 1,2

Note:
If you update the mksysb name for your system partition, you must reapply the system partition configuration using the spapply_config command. For more information on system partitioning, refer to the "Managing system partitions" chapter in PSSP: Administration Guide.

Step 5: Verify installation settings

Make sure that the SDR has the appropriate values specified in the following attributes for each of the nodes. Issue the following command to display the values:

splstdata -G -b

Make sure that /tftpboot/script.cust and /tftpboot/firstboot.cust have been properly updated for |PSSP 3.4 modifications. See Appendix E, User-supplied node customization scripts for additional information.

|Step 6: Refresh RSCT subsystems

The SDR has now been updated to reflect the new nodes that will run |PSSP 3.4. You now need to refresh the |RSCT subsystems on the control workstation and all nodes to pick up these changes. Run syspar_ctrl on the control workstation to refresh the subsystems on both the control workstation and on the nodes.

syspar_ctrl -r -G
Note:
You can ignore any messages that you receive from the nodes you are migrating at this point because the migration process is not yet complete. Once the process is complete, you should no longer receive error messages.

Step 7: Run setup_server to configure the changes

The setup_server command must be run to properly set up NIM on the control workstation by issuing the following command:

setup_server 2>&1 | tee /tmp/setup_server.out

The output will be saved in a log file called setup_server.out.

If you have a node defined as a boot/install server you must also run setup_server out on that node.

dsh -w boot/install_node "/usr/lpp/ssp/bin/setup_server \
    2>&1" | tee /tmp/setup_server.boot/install_node.out

Step 8: Disable nodes from the switch

If you do not have a switch in your SP system, skip this step.

If you want to bring the switch down for all nodes, issue the Equiesce command for each system partition.

Note:
|If you are using the Switch Admin daemon for node recovery, stop it |by issuing stopsrc -s swtadmd on SP Switch systems or stopsrc -s |swtadmd2 on SP Switch2 systems before issuing the Equiesce |command. |

If you use the Equiesce command, you will need to later restart the switch using the Estart command. Issue the Estart command prior to the step where you "Verify the nodes."

If you are migrating a few nodes, you must disable these nodes from the switch (if appropriate, first reassign the primary node or primary backup node). To determine if one of the nodes you are migrating is a primary or primary backup node, issue the Eprimary command. If you need to reassign the primary or primary backup node, issue the Eprimary command with appropriate options. Then issue the Estart command to make your choices effective. You must then issue the Efence command to disable the nodes you are migrating from the switch.

Efence -G node_number node_number

|Step 9: Copy the PSSP 3.4 pssp_script to nodes' /tmp

Copy the |PSSP 3.4 version of pssp_script from the control workstation to the /tmp directory on each of the nodes you are migrating. For example:

pcp -w node /spdata/sys1/install/pssp/pssp_script /tmp/pssp_script

where node is the host name of one or more nodes.

Step 10: Execute the pssp_script on the node

Execute the pssp_script that you copied to /tmp on all of the nodes you are migrating:

dsh -w node /tmp/pssp_script

You need to wait until the script has completed before proceeding to the next step. To determine if pssp_script is still running on the nodes, monitor the 3 digit LEDs and wait for them to turn blank. Check the bootp_response and ensure that it is set to disk by issuing the following command:

splstdata -G -b

You must also update any optional PSSP file sets that you have installed on the node.

Step 11: Reboot the node

If your system contains a switch, you must reboot the node at this time so that changes to the kernel can take affect. Any other kernel change that may have occurred when you upgraded AIX would also require that you reboot the node at this time.

IBM suggests that you reboot the node at this time. To do this, issue:

|cshutdown -rGFN node_list

Step 12: Rejoin the nodes to the switch network

If you disabled all nodes in Step 8: Disable nodes from the switch using the Equiesce command, you must now issue the Estart command in each system partition to rejoin the nodes to the current switch network.

Note:
|If you are using the Switch Admin daemon for node recovery, start it |by issuing startsrc -s swtadmd on SP Switch systems or startsrc |-s swtadmd2 on SP Switch2 systems before issuing the Estart |command. |

If you disabled only a few nodes using the Efence command, you must now issue the Eunfence command to bring those nodes back to the switch network.

|Step 13: Start RSCT subsystems

If you rebooted your node in |Step 11: Reboot the node, you can skip this step because this step would have been automatically performed when the node was rebooted.

To start the subsystems, do the following:

  1. Remove the old subsystems by issuing:
    dsh -w node /usr/lpp/ssp/bin/syspar_ctrl -c
    
  2. Add and start the new subsystems by issuing:
    dsh -w node /usr/lpp/ssp/bin/syspar_ctrl -A
    

To verify that the subsystems are running, refer to Step 56: Verify that RSCT subsystems have started.

Step 14: Run verification tests

Verify that the nodes are running properly by issuing the following commands:

SYSMAN_test
CSS_test              * run only if you have a switch
spverify_config       * run only if your system is partitioned
st_verify             * run only if Job Switch Resource Table Services
                        is installed

Verify that host_responds and switch_responds, if you have a switch, are set to yes by issuing the following command:

spmon -d -G
Note:
If you are migrating nodes in more than one system partition, you need to run CSS_test in each of the system partitions.

|Step 15: Apply PSSP PTFs to nodes (optional)

| |

|Software maintenance (PTFs) may now be applied to the PSSP file sets |installed on the nodes. If you installed software maintenance on the |control workstation before the node migration, you must install the service |updates to the nodes now. Refer to Installing updates on a per node basis for detailed instructions.

|Note:
At this point, refer back to High-level migration steps to determine your next step in the migration |process. |

|


[ Top of Page | Previous Page | Next Page | Table of Contents | Index ]