If you are upgrading from GPFS , you can choose between the following upgrade options. GPFS Administration and Programming Reference Click below to view this publication ?. File System V documentation, where you can find information about how to install, administer, maintain, and use GPFS
|Published (Last):||6 February 2004|
|PDF File Size:||1.22 Mb|
|ePub File Size:||16.95 Mb|
|Price:||Free* [*Free Regsitration Required]|
Offline upgrade, complete cluster 35 This scenario is applicable if you can shut down the entire cluster and perform the upgrade. Online upgrade, most functionality remains available Upgrading to the 5.
Perform any necessary operating system upgrades. At this point, all code is at the upgraded level, but the cluster is not upgraded and new functions cannot be configured until the upgrade is completed. For this upgrade path, use the following platform-specific commands instead of those listed in Step 4: However, due to the entire cluster being offline, it is possible to jump straight to the latest code 35 instead of hopping to an intermediate level, as required during an online upgrade.
Run the necessary platform-specific commands to upgrade GPFS. At this point, you can use the installation toolkit for installation or deployment of new functions and for future upgrades. Re-enable the 35 autoload setting by issuing the following command. At this point, the code is at the upgraded level, but the cluster is not upgraded and new functions cannot be configured until the upgrade is completed.
For each node in the cluster, do the following steps.
Upgrading to IBM Spectrum Scale V from GPFS
You can also create this file system from multiple NSDs, each of which are replicated to help with redundancy. Given the age of GPFS 3. Repeat the preceding steps on each node in the cluster one by one after restarting GPFS.
Alternatively, you can perform function additions and upgrades manually. After the operating system is upgraded and any required reboots are complete, follow the steps for upgrading GPFS.
General Parallel File System V welcome page
At this point all code is at the upgraded level, but gprs cluster is not upgraded and new functions cannot be configured until the upgrade is completed.
For GPFS-only configurations, the process is now complete. Unmount GPFS file systems on the node by issuing the following command.
The steps in this procedure are similar to the steps in the online upgrade procedure. Extract the IBM Gpfe Scale installation image so that all installation files rpm, deb, bff, msi, and so on are available. Confirm that the upgraded operating system level is one that is supported by the version of IBM Spectrum Scale to which the system is being upgraded.
For more information, see Setting up Cluster Export Services shared root file system.
Operating system upgrades while upgrading from GPFS 3. NodeName is the name of the gpfz on which the operating system upgrade will be done.
After all the nodes in the cluster are upgraded to 4. Upgrading from GPFS 3. Regardless of which method you are using, if you want to install and configure CES protocols after upgrading to the latest level, you must create a CES shared root file system of size 4 GB or greater. This file system is used for CES protocol state data and it must remain mounted and up on all protocol nodes for the protocols to function.