Fix Readme
Abstract
xxx
Content
Readme file for: Spectrum Scale
Product/Component Release: 5.0.3.0
Update Name: Spectrum_Scale_Data_Management-5.0.3.0-x86_64-Linux
Fix ID: Spectrum_Scale_Data_Management-5.0.3.0-x86_64-Linux
Publication Date: 19 April 2019
Last modified date: 19 April 2019
Installation information
Download location
Below is a list of components, platforms, and file names that apply to this Readme file.
| Product/Component Name: | Platform: | Fix: |
|---|---|---|
| IBM Spectrum Scale | Linux 64-bit,x86_64 RHEL Linux 64-bit,x86_64 SLES Linux 64-bit,x86_64 Ubuntu | Spectrum_Scale_Data_Management-5.0.3.0-x86_64-Linux |
Prerequisites and co-requisites
Installation information
-
Downloading Images Choose the download option "Download using Download Director" to download the new Spectrum Scale package and place it in any location desired on the install node.
Note, if you must (not recommended) use download option "Download using your browser (HTTPS)", instead of clicking on the down arrow to the left of the package name, you must right-click on the package name and select the Save Link As.. option. If you just click on the download arrow, the browser will likely hang.
-
Installing IBM Spectrum Scale update for System x After you have downloaded the IBM Spectrum Scale 5.0.3.0 update, follow the steps below to install the fix package:
- Ensure the package is executable using ls -l command.
You should see something similar with permissions like:
-rwx r--r-- l root root 110885866 Apr 27 15:52 /download_dir/package_name.
If it's not executable, you can always make the package executable using the following command:
chmod +x /download_dir/package_name - Extract RPMs and Debian Linux packages from Self Extracting Package downloaded using following commands:
For Standard Edition:
./Spectrum_Scale_Standard-5.0.3.0-x86_64-Linux-install
For Advanced Edition:
./Spectrum_Scale_Advanced-5.0.3.0-x86_64-Linux-install
For Data Management Edition:
./Spectrum_Scale_Data_Management-5.0.3.0-x86_64-Linux-install
For Data Access Edition:
./Spectrum_Scale_Data_Access-5.0.3.0-x86_64-Linux-install
Optional Package for SLES and RedHat Enterprise Linux:
- gpfs.docs-5.0.3-0.noarch.rpm
- gpfs.gss.pmcollector-5.0.3-0.xxx.*.rpm (where xxx is the OS version)
- gpfs.gss.pmsensors-5.0.3-0.xxx.*.rpm (where xxx is the OS version)
- gpfs.gui-5.0.3-0.noarch.rpm
- gpfs.java-5.0.3-0.*.rpm
- gpfs.callhome-5.0.3-0.xxx.noarch.rpm (where xxx is the OS type)
- gpfs.callhome-ecc-client-5.0.3-0.noarch.rpm
- gpfs.kafka-5.0.3-0.*.rpm (x86_64 only)
- gpfs.librdkafka-5.0.3-0.*.rpm (x86_64 only)
- gpfs.hdfs-protocol-2.7.3-4.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.hdfs-protocol-3.0.0-0.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.hdfs-protocol-3.1.0-0.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.tct.client-1.1.7*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 and ppc64le only)
- gpfs.tct.server-1.1.7*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 and ppc64le only)
Optional Package for Ubuntu Linux:
- gpfs.docs_5.0.3-0_all.deb
- gpfs.gui_5.0.3-0_all.deb
- gpfs.java_5.0.3-0_*.deb
- gpfs.callhome_5.0.3-0_all.deb (x86_64 only)
- gpfs.callhome-ecc-client_5.0.3-0_all.deb (x86_64 only)
- gpfs.kafka_5.0.3-0_*.deb (x86_64 only)
- gpfs.librdkafka_5.0.3-0_*.deb (x86_64 only)
- gpfs.gss.pmcollector_5.0.3-0.xxx_*.deb (where xxx is the OS version)
- gpfs.gss.pmsensors_5.0.3-0.xxx_*.deb (where xxx is the OS version)
- gpfs.tct.client-1.1.7*.deb (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 only)
- Follow the installation and migration instructions in your IBM Spectrum Scale Installing and upgrading.
- Ensure the package is executable using ls -l command.
-
Upgrading GPFS nodes In the below instructions, node-by-node upgrade cannot be used to migrate from GPFS 4.1 or prior releases. For example, upgrading from 4.1.1.16 to 5.0.3.0 requires complete cluster shutdown, upgrade install on all nodes and then cluster startup.
Upgrading GPFS may be accomplished by either upgrading one node in the cluster at a time or by upgrading all nodes in the cluster at once. When upgrading GPFS one node at a time, the below steps are performed on each node in the cluster in a sequential manner. When upgrading the entire cluster at once, GPFS must be shutdown on all nodes in the cluster prior to upgrading.
When upgrading nodes one at a time, you may need to plan the order of nodes to upgrade. Verify that stopping each particular machine does not cause quorum to be lost or that an NSD server might be the last server for some disks. Upgrade the quorum and manager nodes first. When upgrading the quorum nodes, upgrade the cluster manager last to avoid unnecessary cluster failover and election of new cluster managers.
- Prior to upgrading GPFS on a node, all applications that depend on GPFS (e.g. DB2) must be stopped. Any GPFS file systems that are NFS exported must be unexported prior to unmounting GPFS file systems.
- Stop GPFS on the node. Verify that the GPFS daemon has terminated and that the kernel extensions have been unloaded (mmfsenv -u). If the command mmfsenv -u reports that it cannot unload the kernel extensions because they are "busy", then the install can proceed, but the node must be rebooted after the install. By "busy" this means that some process has a "current directory" in some GPFS filesystem directory or has an open file descriptor. The freeware program lsof can identify the process and the process can then be killed. Retry mmfsenv -u and if that succeeds then a reboot of the node can be avoided.
- Upgrade GPFS as follows(make sure to be in the same directory as the files):
- For Linux:
For SLES or RedHat Enterprise Linux:
rpm -Fvh gpfs*.rpm
rpm -ivh gpfs.license.*.rpm (if you are updating from 4.2.1.2 or older version)
rpm -ivh gpfs.compression.*.rpm
For Debian and Ubuntu Linux:
dpkg -i gpfs*.deb
Recompile any GPFS portability layer modules you may have previously compiled. For more information, reference: Building the GPFS portability layer on Linux nodes
- For AIX:
Use the 'inutoc .' command to create a .toc file which will be used by the installp command. The .toc file will be created in the current working directory.
Once the .toc file is created, Upgrade GPFS using the installp command or via SMIT on the node. If you are in the same directory as the install packages and the .toc file, an example command might be:
installp -agXYd . gpfs
- For Linux:
Additional information
-
Package information The images listed below and contained in the Self Extracting Package (SE-Package) are maintenance packages for IBM Spectrum Scale. The images are a mix of normal RPM or DEB images that can be directly applied to your system.
The packages can be used for new install or update from a prior level of IBM Spectrum Scale.
After all RPMs or DEBs are installed, you have successfully updated your IBM Spectrum Scale product.
Before installing IBM Spectrum Scale, it is necessary to verify that you have the correct levels of the prerequisite software installed on each node in the cluster. If the correct level of prerequisite software is not installed, see the appropriate installation manual before proceeding with your IBM Spectrum Scale installation.
For the most up-to-date list of prerequisite software, see the IBM Spectrum Scale FAQ in the IBM® Knowledge Center .
Update to Version:
5.0.3.0
Update from Version:
4.2.0.0 - 5.0.2.3 (If upgrading node by node )
3.5.0 - 5.0.2.3 (If you shutdown and upgrade the entire cluster)
SE Package Content (SLES and RHEL Linux):
- gpfs.msg.en_US-5.0.3-0.noarch.rpm
- gpfs.base-5.0.3-0.*.rpm
- gpfs.gpl-5.0.3-0.noarch.rpm
- gpfs.docs-5.0.3-0.noarch.rpm
- gpfs.compression-5.0.3-0.*.rpm
- gpfs.gskit-8.0.50-86.*.rpm
- gpfs.gui-5.0.3-0.noarch.rpm
- gpfs.hdfs-protocol-2.7.3-4.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.hdfs-protocol-3.0.0-0.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.hdfs-protocol-3.1.0-0.*.rpm (x86_64, ppc64, and ppc64le only)
- gpfs.java-5.0.3-0.*.rpm
- gpfs.license.xxx-5.0.3-0.*.rpm (where xxx is the license type)
- gpfs.callhome-5.0.3-0.xxx.noarch.rpm (where xxx is the OS type)
- gpfs.callhome-ecc-client-5.0.3-0.noarch.rpm
- gpfs.gss.pmcollector-5.0.3-0.xxx.*.rpm (where xxx is the OS type)
- gpfs.gss.pmsensors-5.0.3-0.xxx.*.rpm (where xxx is the OS type)
- gpfs.kafka-5.0.3-0.*.rpm (x86_64 only)
- gpfs.librdkafka-5.0.3-0.*.rpm (x86_64 only)
- gpfs.adv-5.0.3-0.*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only)
- gpfs.crypto-5.0.3-0.*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only)
- gpfs.tct.client-1.1.7*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 and ppc64le only)
- gpfs.tct.server-1.1.7*.rpm (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 and ppc64le only)
SE Package Content (Ubunut Linux):
- gpfs.msg.en_us_5.0.3-0_all.deb
- gpfs.base_5.0.3-0_*.deb
- gpfs.gpl_5.0.3-0_all.deb
- gpfs.docs_5.0.3-0_all.deb
- gpfs.compression_5.0.3-0_*.deb
- gpfs.gskit_8.0.50-86.*.deb
- gpfs.gui_5.0.3-0_all.deb
- gpfs.java_5.0.3-0_*.deb
- gpfs.kafka_5.0.3-0_*.deb (x86_64 only)
- gpfs.librdkafka_5.0.3-0_*.deb (x86_64 only)
- gpfs.license.xxx_5.0.3-0_*.deb (where xxx is the license type)
- gpfs.gss.pmcollector_5.0.3-0.xxx_*.deb (where xxx is the license type)
- gpfs.gss.pmsensors_5.0.3-0.xxx_*.deb (where xxx is the license type)
- gpfs.adv_5.0.3-0_*.deb (IBM Spectrum Scale Advanced or Data Management Edition only)
- gpfs.tct.client-1.1.7*.deb (IBM Spectrum Scale Advanced or Data Management Edition only, x86_64 only)
-
Summary of changes for IBM Spectrum Scale Unless specifically noted otherwise, this history of problems fixed for IBM Spectrum Scale 5.0.x applies for all supported platforms.
Changes to the IBM Spectrum Scale 5.0.3.0 include the following:
- AFM and AFM DR-related changes
Added a new configuration parameter - afmRefreshAsync to enable cache data refresh operation in asynchronous mod
- Samba
In Spectrum Scale 5.0.3.0 the handling of the NFSv4 ACE bit "SYNCHRONIZE" is changing. When returning ACL data to the SMB client the "SYNCHRONIZE" bit on all allow ACEs will be passed unchanged. Prior to 5.0.3.0 "SYNCHRONIZE" has been automatically set on all allow ACEs by the SMB component, regardless if this bit was set or cleared on the ACLs stored in the file system. Microsoft Windows clients require "SYNCHRONIZE" to be set for renaming files or directories. All files written by Microsoft Windows clients have "SYNCHRONIZE" automatically set. If the ACLs stored in the file system are missing the "SYNCHRONIZE" bit and hence causing access issues with Microsoft Windows clients, restoring the pre 5.0.3.0 behavior is possible by performing the following procedure: for each SMB share requiring the old behavior execute   /usr/lpp/mmfs/bin/net conf setparm
'nfs4:set synchronize' yes and replace with the share name. For the long term it is recommended to change the ACLs for all files or directories missing the "SYNCHRONIZE" bit instead of modifying the SMB configuration. - Cloud services changes
Auto container spillover feature Client-assisted recalls Microsoft Azure object storage service
- File system core improvements
Compatibility settings control the unit types that are returned by statfs and statvfs. The linuxStatfsUnits attribute controls the types of units that are returned in f_bsize, f_rsize, f_blocks, and f_bfree for compatility with earlier or later versions of statfs and statvfs. For more information, see mmchconfig command. zfast, alphae, and alphah libraries compress genomic data File compression methods for genomic data include zfast compression for FASTA and VCF formats and alphae and alphah compression for FASTQ formatted files. The mmkeyserv command displays certificate expiration dates The server show and client show parameters display expiration dates for the REST certificate and the KMIP certificate. The mmnetverify command checks RDMA connectivity The rdma-connectivity option verifies the verbsPorts configuration and checks node connectivity through the active InfiniBand interface. The mmsdrrestore command recovers the CCR state in broken clusters As an alternative to the manual procedures, the --ccr-repair option recovers the CCR state on quorum nodes when CCR files have been lost or corrupted. Other improvements You can use the --use-stale-replica option for the mmfsck command to read possible stale replica blocks from unrecovered disks. You can build the installable package gpfs.gplbin without stopping and restarting GPFS. I/O error checking is improved for locally attached disks of NSD servers. The attribute forceLogWriteOnFdatasync is set to False and can no longer be changed at the user level.
- IBM Spectrum Scale management API changes
Added the following API commands: â¢GET: /filesystems/{filesystemName}/filesets/{filesetName}/quotadefaults â¢POST: /filesystems/{filesystemName}/filesets/{filesetName}/quotadefaults â¢PUT: filesystems/{filesystemName}/filesets/{filesetName}/quotadefaults â¢PUT: /filesystems/{filesystemName}/mount â¢PUT: /filesystems/{filesystemName}/unmount â¢GET: /filesystems/{filesystemName}/quotadefaults â¢POST: /filesystems/{filesystemName}/quotadefaults â¢PUT: /filesystems/{filesystemName}/quotadefaults â¢GET: /filesystems/{filesystemName}/quotagracedefaults â¢POST: /filesystems/{filesystemName}/quotagracedefaults - IBM Spectrum Scale GUI changes
Improved quota management features. The changes made to the Files > Quotas page are as follows: You can now define both capacity and inode quotas for users, user groups, and filesets from the Files > Quotas page. Introduced Defaults tab to improve user experience for defining and enabling default quotas for user, group, and fileset. Introduced Settings tab. You can perform actions such as enable quota, disable quota, set quota accounting scope, check quota, and set grace time periods. For more information on the options that are available in the Quotas page, see Quotas. Introduced Daily quota reports option in the Create Recipient dialog of the Monitoring > Event Notifications page. The recipient can be configured to get reports if soft quota or hard quota is exceeded. You can expand a file systems storage capacity by adding more NSDs to it. Similarly, deleting NSDs from the file system reduces the capacity of the file system. Use the Add NSDs or Remove NSDs options that are available in the Actions menu of the Files > File Systems page to expand or shrink a file system. Added options to configure user authentication for NFS and SMB protocol user access from the Services > File Authentication page. For more information, see File authentication. To prevent files from being changed or deleted unexpectedly, you can use the integrated archive mode (IAM) mode. You can set this option from the Files > Filesets page while creating the fileset or when you modify it. You can modify the list of file auditing events to be included in the file audit log by using the Edit File Audit Logging option that is available in the Actions menu of the Files > File Systems page. Use the Files > Information Lifecycle to define a migration to external pool rule to migrate the data and metadata stored in the local pool to external pools like tape or transparent cloud tiering. Use the Services > GUI > External Authentication option to configure an LDAP-based authentication method on an external authentication server for GUI user access management. NFS and SMB Performance details can be easily monitored through the newly introduced performance charts and metrics in the NFS Exports and SMB Shares pages. The Protocols > NFS Exports page now offers effective management of NFS client definitions and support for pseudo paths for NFSv4. The following navigation menu updates are made in the GUI: Removed the Access and Settings menus. Introduced Cluster and Support menus. Moved File System ACL page from Access menu to Files menu. Moved Remote Connections page from Access menu to newly introduced Cluster menu. Moved Command Audit Log page from the Access menu to Monitoring menu. Moved Event Notifications page from the Settings menu to newly introduced Support menu. Grouped Diagnostic Data and Call Home pages under the new Support menu. Moved Network page to cluster menu.
- Installation toolkit changes
Support for SLES 15 on x86_64 and s390x â¢Significant changes in upgrade process flow: â¦Changes to minimize I/O disruptions â¦Upgrade support for mixed OS cluster â¦Changes in upgrade pre-check to identify the packages that need to be upgraded across the cluster on all nodes. Compare the versions of the installed packages with the versions in the repository of the packages you want to upgrade to. In a mixed operating system cluster, the comparison is done with the package repository applicable for the operating system running on the respective nodes. â¦Changes in upgrade post-check to identify the packages that have not been upgraded successfully â¦Improved dependency check to ensure dependencies are met for each required package
- Object changes
Updated the mmobj command with new password keyword.
- IBM Spectrum Scale SMB changes
Support for SMB 3.1.1. SMB protocol version 3.1.1 is now supported with IBM Spectrum Scale. Support is automatically enabled on new installations. When upgrading, the support can be enabled by making the config change: /usr/lpp/mmfs/bin/net conf setparm global 'server max protocol' SMB3_11 Enhanced the handling of the SYNCHRONIZE bit in NFSv4 ACLs. For more information, see NFS V4 ACL Syntax.
- Watch folder
In IBM Spectrum Scale 5.0.3, watch folder has a clustered watch capability. An entire file system, fileset, or inode space can be watched.
- New subroutines
gpfs_add_filesystem_watch.
- Changed commands
â¢mmbuildgpl â¢mmchattr â¢mmchconfig â¢mmchfileset â¢mmchfs â¢mmcrfs â¢mmfsck â¢mmkeyserv â¢mmlsattr â¢mmnetverify â¢mmobj â¢mmprotocoltrace â¢mmsdrrestore â¢mmwatch
- AFM and AFM DR-related changes
Was this topic helpful?
Document Information
Modified date:
03 May 2019
UID
isg400004417
