About this document
This paper is intended as an architecture and configuration guide to set up the IBM® System Storage® for the SAP HANA tailored data center integration (SAP HANA TDI) within a storage area network (SAN) environment. SAP HANA TDI allows the SAP customer to attach external storage to the SAP HANA server.
The paper also describes the setup and configuration of SAP Landscape Management for SAP HANA systems on IBM infrastructure components: IBM Power Systems™ and IBM Storage based on IBM Spectrum™ Virtualize.
This document is written for IT technical specialists and architects with advanced skill levels on SUSE Linux Enterprise Server or Red Hat Enterprise Linux (RHEL) and IBM System Storage.
This document provides the necessary information to select, verify, and connect IBM System Storage to the SAP HANA server through a Fibre Channel-based SAN. The recommendations in this Blueprint apply to single-node and scale-out configurations, and Intel and IBM Power based SAP HANA systems.
For more information about the storage systems that are certified for SAP HANA production, see the Certified and Supported SAP HANA Hardware Directory.
Scope
This Blueprint provides a solutions architecture and the related storage endpoint capabilities that interact with the following software and hardware components:
•IBM Spectrum Virtualize™
•IBM SAN Volume Controller
•IBM FlashSystem Family
•IBM Spectrum Protect™
•IBM Spectrum Protect for Enterprise Resource Planning
•IBM Geographically Dispersed Resiliency for IBM Power Systems
•Linux operating system (SLES or RHEL)
•SAP Landscape Management
•IBM Power Systems
•IBM Power Virtualization Center (PowerVC)
This technical report does not:
•Provide performance analysis from a user perspective
•Replace any official manuals and documents issued by IBM
•Explain installation and configuration of SAP HANA or SAP S/4HANA
Prerequisites
This technical paper relegates to SAP documentation that is available for download from the SAP websites:
It is assumed that you are familiar with and have basic knowledge of the following products:
•IBM FlashSystem
•IBM Spectrum Protect
•IBM Spectrum Protect for Enterprise Resource Planning
•Linux operating system (SLES or RHEL)
•SANs
Foundation: Data Placement and Data Management
This section covers data placement and data management.
IBM storage architecture for SAP HANA TDI Environment
IBM System Storage must be connected through a SAN to the SAP HANA server. IBM recommends using IBM SAN Volume Controller as a virtualization layer to enable centralized management of the storage systems and easy high-availability and disaster-recovery functionality, and IBM FlashCopy® backup and tiered-storage pooling.
Figure 1 IBM Spectrum Virtualize architecture landscape diagram
Figure 1 shows the recommended architecture for SAP, with IBM SAN Volume Controller configured as an enhanced stretched cluster. All data, which is stored on IBM SAN Volume Controller as virtual volumes, is mirrored through IBM SAN Volume Controller virtual disk mirroring onto the physical storage system on sites A and B. These storage systems are a mix of IBM FlashSystem hardware, and are configured depending on the workload, based on the total number of SAP instances and the size of each SAP instance.
To set up a stretched storage cluster, a tiebreaker (quorum) mechanism is required. A requirement met in this case with an IP quorum.
Implementation requirements
Identify the appropriate size for your IBM Spectrum Virtualize environment, and then review requirements for hardware and software based on your SAP HANA landscape.
Use Table 1 to determine the capacity of the IBM Storage systems that you need, based on the number of SAP instances you manage. Both the total number of SAP instances and the database size in terabytes (TB) are shown in Table 1.
Table 1 Storage sizing for SAP HANA
|
Number of SAP HANA
instances (max)
|
Size of each SAP HANA
instance on disk (max)
|
IBM Storage system
sizing
|
|
5 Prod + 10 Dev/Test
|
0.5 TB
|
Small
|
|
10 Prod + 20 Dev/Test
|
1 TB
|
Medium
|
|
10 Prod + 20 Dev/Test
|
2 TB
|
Large
|
|
10 Prod + 20 Dev/Test
5 Prod
|
2 TB
4 TB
|
Extra Large
|
This table guides you through your storage sizing. If you need more precise sizing, ask your IBM sales representative or IBM Business Partner for a sizing with the IBM HANAmagic tool, which is available to IBM sales team members and IBM Business Partners only.
Hardware requirements
This section covers the hardware requirements for small, medium, large, and extra large configurations. Table 2 on page 4, Table 3 on page 5, Table 4 on page 6, and Table 5 on page 7 contain the descriptions and quantities for IBM components that are used in the storage-configuration Blueprints.
The configurations are recommendations to meet the SAP HANA TDI performance and capacity specifications determined by SAP Quick Sizer.
The hardware specifications that are provided are current at the time of this writing. Part substitutions might be required, depending on hardware availability. Be cautious if you plan to substitute a smaller quantity of higher-capacity drives, particularly for the database. A smaller quantity of drives might not provide comparable performance.
The tables contain abbreviated part lists, which include only major components. Work with your hardware supplier to ensure that your configuration is complete.
The tables include IBM SAN Volume Controller and IBM FlashSystem to provide for an enhanced cluster setup, which is a recommended high availability option for SAP HANA landscapes.
You must provide or acquire hardware that is based on appropriate scale requirements. You can choose components that are equivalent to or better than those listed.
|
Note: Part numbers for IBM warranties are not included here. There are standard offerings for 1-year and 3-year warranties. The warranties that are available for purchase depend on your configuration requirements.
|
A SAN switch must be included, with enough ports to match those of the physical SAP HANA server.
A SAN environment likely exists and might need to be extended.
Hardware requirements for small-sized landscapes
Entry enterprise IBM FlashSystem solutions are designed to provide simple, high-performance management of block-based storage. For this reason, IBM FlashSystems are suited for use by SAP HANA. The configuration is based on Table 2.
Table 2 Hardware requirements for small-sized landscapes
|
Blueprint
component
|
Description
|
Quantity
|
Part number
|
|
IBM SAN Volume Controller
|
IBM SAN Volume Controller Storage Engine
|
|
2147-SV1
|
|
|
IBM SAN Volume Controller Initial Storage Engine
|
|
3001
|
|
|
Additional IBM SAN Volume Controller Storage Engine
|
|
3002
|
|
|
4-port, 16 Gb Fibre Channel card,
per engine |
|
AH14
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller V8 |
|
5641-VC8
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller 3Y |
|
5641-VCC
|
|
|
IBM Enterprise Class Support for IBM SAN Volume Controller
|
|
5773-ECS
|
|
IBM Storage
|
IBM FlashSystem 5030 Controller
|
|
2072-3H4
|
|
|
16 Gb Fibre Channel adapter pair
|
|
ALBB
|
|
|
Controller cache upgrade
|
|
ALGA
|
|
|
7.68 TB, 2.5-inch
flash drive (solid state drive [SSD]) |
12
|
AL82
|
|
IP quorum
|
Linux system, native or as a
virtual machine (VM) |
|
RHEL > V7.1
SLES > V11.3 |
Hardware requirements for medium-sized landscapes
Table 3 lists the hardware requirements for medium-sized landscapes.
Table 3 Hardware requirements for medium-sized landscapes
|
Blueprint
component
|
Description
|
Quantity
|
Part number
|
|
IBM SAN Volume Controller
|
IBM SAN Volume Controller Storage Engine
|
|
2147-SV1
|
|
|
Initial IBM SAN Volume Controller Storage Engine
|
|
3001
|
|
|
Additional IBM SAN Volume Controller Storage Engine
|
|
3002
|
|
|
4-port, 16 Gb Fibre Channel card,
per engine |
|
AH14
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller V8 |
|
5641-VC8
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller 3Y |
|
5641-VCC
|
|
|
IBM Enterprise Class Support for IBM SAN Volume Controller
|
|
5773-ECS
|
|
IBM Spectrum Storage™
|
IBM FlashSystem 5030 Controller
|
|
2072-3H4
|
|
|
16 Gb Fibre Channel adapter pair
|
|
ALBB
|
|
|
Controller cache upgrade
|
|
ALGA
|
|
|
7.68 TB, 2.5-inch flash drive (SSD)
|
24
|
AL82
|
|
IP quorum
|
Linux system native, or as a VM
|
|
RHEL > V7.1
SLES > V11.3 |
Hardware requirements for large-sized landscapes
Table 4 lists the hardware requirements for large-sized landscapes.
Table 4 Hardware requirements for large-sized landscapes
|
Blueprint
component
|
Description
|
Quantity
|
Part number
|
|
IBM SAN Volume Controller
|
IBM SAN Volume Controller Storage Engine
|
|
2147-SV1
|
|
|
Initial IBM SAN Volume Controller Storage Engine
|
|
3001
|
|
|
Additional IBM SAN Volume Controller Storage Engine
|
|
3002
|
|
|
4-port, 16 Gb Fibre Channel card,
per engine |
|
AH14
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller V8 |
|
5641-VC8
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller 3Y |
|
5641-VCC
|
|
|
IBM Enterprise Class Support for IBM SAN Volume Controller
|
|
5773-ECS
|
|
IBM Spectrum Storage
|
IBM FlashSystem 5030 Controller
|
|
2072-3H4
|
|
|
16 Gb Fibre Channel adapter pair
|
|
ALBB
|
|
|
Controller cache upgrade
|
|
ALGA
|
|
|
7.68 TB, 2.5-inch flash drive (SSD)
|
24
|
AL82
|
|
IP quorum
|
Linux system, native or as a VM
|
|
RHEL > V7.1
SLES > V11.3 |
Hardware requirements for extra large-sized landscapes
For even larger SAP HANA deployments, IBM FlashSystem solutions are a good fit with respect to maximum performance and density (TB per rack unit). Also, the configuration for IBM SAN Volume Controller includes spare nodes, for increased availability.
Table 5 lists the hardware requirements for extra large-sized landscapes.
Table 5 Hardware requirements for extra large-sized landscapes
|
Blueprint
component
|
Description
|
Quantity
|
Part number
|
|
IBM SAN Volume Controller
|
IBM SAN Volume Controller Storage Engine
|
|
2147-SV1
|
|
|
Initial IBM SAN Volume Controller Storage Engine
|
|
3001
|
|
|
Additional IBM SAN Volume Controller Storage Engine
|
|
3002
|
|
|
Standby IBM SAN Volume Controller Storage Engine
|
|
3003
|
|
|
4-port, 16 Gb Fibre Channel card,
per engine |
|
AH14
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller V8 |
|
5641-VC8
|
|
|
IBM Spectrum Virtualize software
for IBM SAN Volume Controller 3Y |
|
5641-VCC
|
|
|
IBM Enterprise Class Support for IBM SAN Volume Controller
|
|
5773-ECS
|
|
IBM Spectrum Storage
|
IBM FlashSystem 9110 SFF controller
|
|
9730-AF7
|
|
|
120 GB, M.2 boot drive pair
|
|
ACGW
|
|
|
16 Gb Fibre Channel adapter pair
|
|
AHB3
|
|
|
Base cache
Cache upgrade |
1
1 |
ACG0
ACGA |
|
|
9.6 TB NVM Express (NVMe) IBM FlashCore modules
|
24
|
AHS2
|
|
|
IBM Spectrum Virtualize software
for IBM FlashSystem |
|
5639-FA2
|
|
IP quorum
|
Linux system, native or as a VM
|
|
RHEL > V7.1
SLES > V11.3 |
Physical installation and setup of IBM Storage Systems
This section describes how to set up and configure the storage systems that are outlined. The physical setup of the appropriate IBM storage systems is described in the following documents:
IBM System Storage SAN Volume Controller Models 2145 and 2147 Hardware Installation Guides
Planning and Installing
Configuring the storage landscape
This section describes configuring the IBM storage systems and the Linux operating system in preparation for SAP HANA.
Figure 2 on page 9 shows the following logical storage relationship:
•One file system (here, XFS) is built on one logical volume (LV)
•One or more LVs are in one volume group (VG)
•One VG is placed on one or more physical volumes (PVs, a Linux term)
•One PV is one virtual disk (VD, an IBM term)
•One or more VDs are placed in two pools via virtual disk mirroring
•One pool is built of one or more managed disks (MDs)
•One MD is one VD of an external storage system (here, IBM FlashSystem) or one MD is one distributed array of internal drives (here, SSDs)
Figure 2 Logical storage configuration
The following central schema is used to set up the infrastructure for SAP HANA TDI:
1. Verify that you met all SAP-provided infrastructure requirements.
2. Configure the IBM storage systems.
3. Create VMs or logical partitions (LPARs) for the HANA systems.
4. Create SAN zones between storage systems and VMs or LPARS.
5. On the storage systems, map the storage volumes by logical unit number (LUN) to the target SAP HANA nodes.
6. Install the Linux operating system on VMs.
7. Perform Linux setup and tuning, including multipath and input/output (I/O) scheduler configuration.
8. On each Linux host, create the file systems for SAP HANA data, log, and shared.
9. Set up Network File System (NFS) services for the SAP HANA shared directory if you plan to scale out your HANA deployment by using a solution, such as SAP Business Warehouse (BW).
Applying SAP requirements
Consider the following primary SAP Notes for an IBM landscape:
•2378874: Install SAP Solutions on Linux on IBM Power Systems (little endian)
•2055470: SAP HANA on IBM Power Systems and IBM System Storage Planning and Installation Specifics: Central Note
SAN zoning between IBM SAN Volume Controller and IBM FlashSystem
The SAN administrator needs to create four SAN zones. For more information, see the section “Edge-core-edge topology” of Chapter 1, Storage Area Network, of IBM System Storage SAN Volume Controller, and IBM FlashSystem 7200 Best Practices and Performance Guidelines.
Configuring IBM FlashSystem
All configuration can be completed via the graphical user interface (GUI). This GUI eases the set-up process compared to using the command-line interface (CLI). The IBM GUI guides you through the initial setup, including configuration of a distributed array and distributed RAID (DRAID).
Complete the following steps to configure the IBM FlashSystem to be used by IBM SAN Volume Controller:
1. Create a single pool, called Pool1.
2. Include all SSDs in one distributed array, using DRAID 6, and map that array to Pool1; turn on IBM Easy Tier®; turn off compression.
3. For small or medium configurations, create 4 VDs of equal size, using 80% of the available Pool1 capacity.
4. For large configurations, create 8 VDs of equal size, using 80 percent of the available Pool1 capacity.
5. Map the VDs to IBM SAN Volume Controller.
Configuring IBM SAN Volume Controller
To set up IBM SAN Volume Controller to provide an enhanced stretched cluster, follow the procedures up to Chapter 4 in IBM Spectrum Virtualize and SAN Volume Controller Enhanced Stretched Cluster with VMware.
Installing Linux
Install the Linux operating system according to the SAP Product Availability Matrix (PAM) for the chosen SAP HANA version. For more information, see SAP Note 1944799.
Creation of VMs (or LPARs)
Create VMs via IBM Power Virtualization Center (IBM PowerVC). This setup is not part of this document.
During this setup, IBM PowerVC also configures the appropriate SAN zones.
Creating volumes for SAP HANA VM
Figure 3 shows the required file system layout. Each colored box indicates one mount point with one LV and one VG.
Figure 3 File system layout for SAP HANA VM
Consider the following sizing guidelines:
•Use one 100 GB physical disk (PD) for root VG
•Use one 50 GB PD for /usr/sap/SID VG
•Use one 100 GB PD for /hana/shared/SID VG
•Use four PDs for /hana/data/SID VGs, sized to anticipate growth, according to HANA sizing guidelines
•Use four PDs (maximum size 500 GB) for /hana/log/SID VGs, sized according to HANA sizing guidelines
Creating the file systems for SAP HANA
Complete the following steps to create the file system for SAP HANA:
1. Use four PVs for /hana/data/log and four for /hana/log/SID; use one for each /usr/sap/SID and one for /hana/shared/SID. ANA (ana) is used as SID in this example.
2. Create a VG with an alignment size of 1 MB:
vgcreate -s 1M --dataalignment 1M analogvg /dev/mapper/WWID_1 dev/mapper/WWID_2 ...
3. Create an LV with a stripe size of 256 KB:
lvcreate -i 4 -I 256K -l 100%VG -n analoglv analogvg
4. Create the file system:
mkfs.xfs -b size=4096 -s size=4096 /dev/analogvg/analoglv
Figure 4 File systems
6. Mount the file system:
mount /hana/log/ANA
mount ...
Required Linux tuning
Tune the Linux multipath settings and I/O scheduler selection.
Multipath
Figure 5 shows a sample configuration /etc/multipath.conf, which is tuned for IBM SAN Volume Controller with flash or SSD storage.
Figure 5 Sample configuration
Using user-friendly names
Optionally, you can use user-friendly names, as shown in Figure 6 on page 12, which makes managing and identifying volumes easier.
Figure 6 User friendly naming
I/O scheduler (elevator kernel parameter)
The default Completely Fair Queuing (CFQ) scheduler delivers low I/O performance for multiple parallel streaming read processes. The NOOP scheduler is recommended for setups with devices that do I/O scheduling themselves, such as intelligent storage devices, or in multi-path environments. The Deadline scheduler can provide a superior throughput over the CFQ I/O scheduler in cases where several threads read and write, and fairness is not an issue (for example, for several parallel readers from a SAN).
We recommend using the NOOP scheduler for SAP HANA.
For more information, see SUSE System Analysis and Tuning Guide.
HANA configuration for IBM SAN Volume Controller
The following fileio parameters must be set for IBM storage systems by using the SAP HANA tool hdbparam:
async_write_submit_active : on
async_write_submit_blocks : all
async_read_submit : on
For more information about the SAP HANA tool, see this web page (log in required).
Data protection: Backup and recovery
SAP HANA backups are required to protect against disk failures and to enable earlier database resets. To perform online backups of SAP HANA, you can use the following IBM software tools:
•IBM Spectrum Protect for Enterprise Resource Planning: Data Protection for SAP HANA
•IBM FlashCopy backup for SAP HANA (FCTool)
Figure 7 on page 14 shows the solution architecture of the virtualized SAP HANA environment in combination with Data Protection for SAP HANA, the FlashCopy backup solution, and IBM Spectrum Virtualize which serves as the virtualization layer for storage.
If you experience a total loss of the SAP HANA appliance, use the SAP HANA recovery procedure to recover the system. For more information about hardware recovery, see the appropriate SAP documentation.
The following sections show where to find detailed information about the Data Protection and FlashCopy backup solutions.
Figure 7 Solution architecture
SAP HANA backup, IBM Spectrum Protect for Enterprise Resource Planning
IBM Spectrum Protect, formerly known as IBM Tivoli® Storage Manager for Enterprise Resource Planning, includes the package Data Protection for SAP HANA.
For more information about how to install, configure, and run this integrated solution, see the IBM Spectrum Protect for Enterprise Resource Planning documentation.
IBM FlashCopy backup for SAP HANA
To minimize backup and restore time, IBM offers a FlashCopy solution that combines SAP HANA snapshot technology with IBM Spectrum Virtualize. For more information, see IBM FlashCopy backup and copy solution for SAP HANA TDI, V4.1.
Data placement: SAP data aging
For many years, SAP classified internal data based on its age and access pattern. For more information, see the following SAP documentation:
Dynamic tiering
Dynamic tiering is available for SAP BW on IBM DB2® and Oracle databases. Dynamic tiering distinguishes between:
•Hot: Loaded in HANA random access memory (RAM) during database startup
•Warm: Loaded into RAM on demand
•Cold: Stored outside the primary HANA database, but in a secondary anyDB database or SAP IQ via the nearline storage (NLS) interface
•Frozen: Stored outside any SAP database, or was archived and then deleted from the SAP database
SAP dynamic tiering retains Warm data in a HANA persistence layer and the data is brought into memory only when needed. Infrequently accessed data is retained in the persistence layer; if accessed frequently, it overwhelms the system.
Dynamic tiering can be installed on its own server, but not on its own HANA node. It is provided as an add-on option for native SAP HANA applications and at the time of this writing, SAP is evaluating dynamic tiering as an option for S/4HANA.
Data tiering optimization
Data lifecycle management strategies can be implemented with the data tiering optimization option. Data tiering optimization (DTO) is available for SAP BW on HANA or SAP BW/4HANA only and distinguishes between:
•Hot: Processed and stored on the primary HANA node
•Warm: Processed and stored on extension nodes
•Cold: Processed and stored outside the primary and extension nodes on SAP IQ (not anyDB), and accessed via an NLS interface
DTO improves data management by shifting data to the performance, storage, and processing tier with the cost and performance characteristics that are best suited for that data. Data tiering options from SAP provide a trusted approach for intelligent enterprises to manage data growth effectively and economically.
Data tiering is similar to conventional BW on HANA scale-out, but also provides the benefits of lower CPU/RAM requirements on the extension nodes.
Data aging
Data aging is available only for S/4HANA, and distinguishes between:
•Hot: Loaded in HANA RAM during startup
•Warm: Loaded into RAM on demand
Impact to storage
In all three cases, the hot and warm data is on a file system and stored on a storage system.
Especially with S/4HANA data aging, HANA acts as a classical database, such as IBM DB2. Data is loaded from the file system into RAM on demand, instead of entirely at startup, because the database size (in the file system) is almost always much larger than the available RAM.
This IBM Blueprint fully supports all three SAP aging scenarios, if you decide to set up data aging for your SAP instance.
Data protection: Disaster recovery
IBM Spectrum Protect is the eminent IBM enterprise backup and restore (EBR) tool. A key differentiator for IBM Spectrum Protect is that it performs a progressive incremental forever backup, which means that only a single full backup is required. All subsequent backups contain only incremental changes. This approach means that IBM Spectrum Protect solutions require less storage, compute, and network infrastructure when compared to other EBR solutions.
Through integration between the solutions, IBM Spectrum Protect and IBM Spectrum Virtualize manage and protect the huge volumes of data for big data analytics and new cognitive workloads.
Figure 8 shows the architecture for a possible disaster-recovery scenario. In this use case, we have two on-premises data centers.
Figure 8 Possible disaster-recovery scenario
Site A is the main site. Under normal operations, IBM Spectrum Virtualize manages the storage resources needed by the VMs and their associated database applications, such as SAP S/4HANA. IBM Spectrum Protect is providing backup, archive, recovery, space management and, if necessary, bare-metal system recovery.
Site B is the failover site. It is considered to be on-premises as well. IBM Global Mirror is mirroring the data on the IBM SAN Volume Controller cluster from Site A to Site B. IBM Geographically Dispersed Resiliency for Power Systems serves as the cluster manager that pairs the VMs across the two sites. This pairing ensures that each site has the resources available to run the host if a planned or unplanned move occurs. IBM Geographically Dispersed Resiliency for Power Systems is installed on both sites.
In case of a physical disaster on Site A, HANA can be restarted on Site B. In the case of a logical disaster, the data can be restored from the IBM Spectrum Protect instance on Site A. There is some amount of downtime, but a business can successfully recover from either scenario, whether a physical or a logical disaster.
Disaster recovery with geographically dispersed resiliency
IBM Geographically Dispersed Resiliency for Power Systems is a set of software components that provide a disaster-recovery mechanism for VMs that are running on an IBM POWER® processor-based server.
The main purpose of disaster recovery is to have a defined (and possibly automated) procedure for recovery from a major business impact, such as an outage of the entire data center as the result of an earthquake, flood, or storm.
Recovery time objective (RTO) and recovery point objective (RPO) for disaster recovery are normally different from the RTO and RPO values for availability.
IBM Geographically Dispersed Resiliency for Power Systems offers the following benefits:
•Simplified disaster-recovery management
•Cost savings based on elimination of the need for hardware and software resources at the backup site
•Reduced licensing costs
•Reduced administrative costs
•Ease of deployment because no high-availability clustering is required
•The ability to test the disaster-recovery environment without affecting the production environment
IBM Geographically Dispersed Resiliency for Power Systems uses VM technology to restart VMs at a backup site if there is a disaster or planned system maintenance by the administrator. All of this is controlled by a single management system that is separate from the production systems. Replication is performed at the storage layer, which provides storage-based mirroring with automated failover.
IBM Geographically Dispersed Resiliency for Power Systems consists of several necessary components (see Figure 8 on page 16) that work together to provide a highly available environment for systems across two sites.
Figure 9 shows the principal setup. To learn about the various components, subsystems, and tasks that are associated with IBM Geographically Dispersed Resiliency for Power Systems, see IBM Geographically Dispersed Resiliency for Power Systems.
Figure 9 Components of IBM Geographically Dispersed Resiliency for Power Systems
SAP Landscape Management
This section covers SAP Landscape Management.
SAP HANA infrastructure
This section covers the technical overview and SAP Landscape view.
Technical overview and SAP Landscape overview
SAP Landscape Management (LaMa) provides functionalities that customers can use to streamline their IT operation for SAP environments, such as:
•Monitoring
•Start, stop, and relocate SAP systems (including mass operations)
•Provisioning of SAP systems, operating systems, virtual machines, and storage volumes (LUNs)
•SAP system copy, cloning, and refresh
Some of these functionalities are provided by SAP, others integrate with SAP partner solutions. The IBM integrated solutions are: IBM PowerVC, IBM Spectrum Protect Snapshot (FlashCopy Manager) (see Figure 10 on page 19).
Figure 10 SAP LaMa infrastructure example
Most relevant for IBM Storage is the system copy (cloning) and the provisioning parts of SAP Landscape Management. These are the use cases we are addressing with this Blueprint.
SAP Landscape Management uses IBM PowerVC or IBM Spectrum Protect Snapshot to initiate FlashCopies or snapshots on IBM storage. IBM PowerVC is also used for the provisioning of LPARs (VMs), storage volumes (LUNs), and so on (see Figure 11).
Figure 11 IBM Power Systems and storage management with IBM PowerVC
In combination with Post-Copy Automation (PCA, delivered by SAP) customers can reduce the time required to set up a QA, test, or education system (as copy from a production system) from days to a few hours. In this case, storage-based snapshots replace time-consuming backups and restores of SAP systems and PCA handles application-specific tasks that are required after a system copy.
The users of LaMa are SAP administrators who are interested in SAP systems, databases, and sufficient file systems capacity. IT operations (such as LUN provisioning and operating system installation) are started by LaMa, and can be transparent to the SAP administrator, but not to the user who configures LaMa. You can compare the roles to the car driver and the mechanic.
Figure 12 shows the minimum landscape to set up and test the LaMa functions in the IBM lab.
Figure 12 SAP Landscape Management 3.0 setup
|
Note: The SAP note 2050537, which is titled “Support for SAP HANA in SAP Landscape Management”, provides a consolidated overview of the SAP Landscape Management release history and associated functionalities.
|
SAP Netweaver installation (and upgrade to latest SP)
An SAP Landscape Management installation requires the following major steps:
•Install SAP NetWeaver Application Server 7.5 for Java.
•Download and install the Landscape Management software packages on the NetWeaver Application Server.
SAP NetWeaver runs on multiple operating systems. We chose IBM AIX® and set up the following software:
•IBM AIX 7.2 Technology Level 03 Service Pack 2 Built 1845
•SAP NetWeaver 7.50
•SAP Landscape Management 3.0 SP10
The prerequisites and the installation process for SAP NetWeaver 7.5 is described in the SAP manual Installation of SAP Systems Based on the Application Server Java of SAP NetWeaver 7.1 to 7.5 on UNIX: IBM DB2 for Linux, UNIX, and Windows (Database: IBM DB2 for Linux, UNIX, and Windows, Operating System: UNIX/Linux).
For more information, see the following SAP Notes:
•1972803: SAP on AIX: Recommendations
•1707361: Inst. Systems Based on NW 7.1 and Higher: UNIX DB2 for LUW
These manuals and SAP notes describe the hardware and software requirements and the file system layout that is required for SAP NetWeaver with a specific database. We chose IBM DB2 database. Table 6 shows a sample file system layout.
Table 6 Sample file system layout
|
Filesystem
|
GB blocks
|
Free
|
%Used
|
Iused
|
%Iused
|
Mounted on
|
|
/dev/db2bigSIDlv
|
19.88
|
19.71
|
1%
|
75
|
1%
|
/db2/ELM
|
|
/dev/db2sidlv
|
5.31
|
1.70
|
69%
|
7256
|
2%
|
/db2/db2elm
|
|
/dev/logdirlv
|
23.88
|
19.20
|
20%
|
67
|
1%
|
/db2/ELM/log_dir
|
|
/dev/sapdatalv
|
49.88
|
44.38
|
12%
|
10
|
1%
|
/db2/ELM/sapdata1
|
|
/dev/saptemp1lv
|
9.88
|
9.83
|
1%
|
4
|
1%
|
/db2/ELM/saptemp1
|
|
/dev/sapmntlv
|
11.88
|
9.19
|
23%
|
1233
|
1%
|
/sapmnt
|
|
/dev/usrsaplv
|
59.38
|
42.48
|
29%
|
550624
|
6%
|
/usr/sap
|
The SAP NetWeaver installation is described in the SAP manual, SAP Landscape Management 3.0, Enterprise Edition, Installation Guide, on the SAP Help Portal.
The installation of SAP NetWeaver requires two SAP tools:
•Software Provisioning Manager (SAP installer, sapinst) to install SAP components
•Software Update Manager 1.0 to perform release upgrades, update single components, and install additional packages
We also recommend to use the SAP Maintenance Planner to find and download appropriate and compatible software packages.
To work with the SAP Maintenance Planner, see the SAP Help Portal and search for “Maintenance Planner”. Then, click the button of the corresponding application.
|
Note: The SAP note 2350235 “SAP Landscape Management 3.0 - Enterprise edition” provides more information about SAP Landscape Management 3.0, enterprise edition, as an add-on to SAP NetWeaver Application Server for Java.
|
SAP Landscape Management installation and upgrade to latest SP
SAP provides the Landscape Management software called SCA (SAP Component Archive) files, which can be installed and updated by using the Software Update Manager (SUM).
SUM and the SCA files are available for download from the SAP Support Portal. At the portal, select Download Software.
Log on to the LaMa server as <sid>adm and unpack the Software Update Manager package by using the following command:
SAPCAR –xf <download directory>/<path>/<Archive>.SAR -R /usr/sap/<sid>
For example:
./sapcar -xf /instimages/SUM/SUM10SP23_1-20006540.SAR -R /usr/sap/ELM
To start the Software Update Manager, run the following command as the root user:
/usr/sap/<SID>/SUM/STARTUP <SID>
Then, open a web browser and access the Software Update Manager by using the following URL:
http://<hostname>:1128/lmsl/sumjava/<SID>/index.html
The Software Update Manager guides you through the installation process (see Figure 13).
Figure 13 SAP LaMa upgrade with Software Update Manager
Three SCA packages are required for the SAP LaMa 3.0 Enterprise Edition: VCM, VCM_ENT, and VCMCR4E.
For more information about SUM and the LaMa installation, see the following SAP manuals:
•Updating SAP Java Systems on UNIX and Linux: IBM DB2 for Linux, UNIX, and Windows (Upgrade Guide -Software Update Manager 1.0)
•SAP Landscape Management 3.0, Enterprise Edition, Installation Guide
Initial SAP Landscape Management setup
Now that SAP LaMa is installed, log on to the SAP NetWeaver Administrator (NWA) application and complete the following steps:
1. Open http://lamahost:50000/nwa URL in a web browser.
2. In the SAP NetWeaver Administrator application, select Configuration → Scenarios → Configuration Wizard. Then, select Initial Setup for SAP Landscape Management, (see Figure 14).
Figure 14 Initial setup for SAP LaMa with the Configuration Wizard in SAP NWA
3. Access the SAP Landscape Management application by using the URL and perform the basic configuration steps for your SAP landscape.
More information
The SAP Landscape Management 3.0, Enterprise Edition User Guide is available in the SAP Help Portal and describes the initial setup and the configuration steps for the LaMa use cases. Access the SAP Help Portal and search for “Landscape Management 3.0” to download this manual.
More information is available in the SAP note 2039615: Managing system landscapes with SAP Landscape Management Enterprise Edition.
The configuration steps that are required for this IBM Blueprint sample setup are described in “SAP LaMa basic setup and configuration” on page 28.
IBM System integration to SAP LaMa
This section describes the integration of SAP LaMa into an IBM Power Systems environment. For the integration of virtualized IBM PowerLinux™ systems and IBM storage the IBM Power Virtualization Center (PowerVC) is required, as shown in Figure 15 on page 24. If the SAP systems was installed on Linux_x86, the PowerVC counterpart IBM Spectrum Protect Snapshot is required.
The IBM components to integrate into SAP LaMa are:
•IBM Power Virtualization Center (IBM PowerVC) for virtualization management
•IBM FlashSystem 9200 storage system
•IBM PowerSystem H922
•SAP Adaptive Extensions with IBM libraries
Figure 15 Integration of SAP LaMa with IBM Power Systems
For the host communication with SAP LaMa, the SAP Host Agent must be installed on every managed SAP host. The SAP Host Agent is the most important agent in SAP LaMa. Ensure that you keep the Host Agent updated with the latest patch level on each host in the landscape. The SAP Host Agent provides an update feature that helps you with that task.
The SAP Host Agent enables SAP LaMa to start and stop instances, and to monitor the status of computer systems.
For more information, see SAP note 1031096: Installing Package SAPHOSTAGENT.
IBM PowerVC Setup
IBM PowerVC Virtualization Center is an advanced virtualization and cloud management offering, which is built on OpenStack, that provides simplified virtualization management and cloud deployments for IBM AIX, IBM i, and Linux VMs running on IBM Power Systems. PowerVC is designed to improve administrator productivity and simplify the cloud management of VMs on Power Systems servers.
To manage SAP HANA environments on PowerLinux with SAP LaMa, PowerVC must be integrated into SAP LaMa. PowerVC manages the IBM Power Systems and IBM storage system for the SAP LaMa tasks.
|
Note: The minimal supported IBM PowerVC version is 1.4.0.
|
To install PowerVC, see Installing IBM Power Virtualization Center.
The pre-installation checklist for PowerVC (PVC_checklist.pdf that is attached to SAP Note 2372251) provides more information about the environment that is required for supporting the IBM PowerVC Storage Adapter for SAP LaMa 3.0. It also specifies the supported versions of the hardware and software components that compose the system landscape for SAP LaMa Tasks.
IBM PowerVC includes the following tasks:
1. An IBM PowerVC user with Administrator privileges is required. SAP LaMa uses this user for creating the LPARs, creating and deleting LUNs, and mapping them to the target LPAR.
2. Communication between SAP LaMa and IBM PowerVC by using “server certificates” provides the most secure level. The IBM PowerVC certificate is used, and the host name within the certificate is verified against the host name in SAP LaMa. In this case, the IBM PowerVC server certificate must be imported into the keystore of the SAP NetWeaver engine on which SAP LaMa is running.
3. PowerVC must be defined in LaMa as Virtualization manager and as Storage manager as described in the following steps:
a. To register PowerVC as Virtualization manager, navigate to Configuration → Infrastructure → Virtualization Managers, as shown in Figure 16.
Figure 16 PowerVC integration as Virtualization manager
b. To register PowerVC as Storage manager, navigate to Configuration → Infrastructure → Storage Managers, as shown in Figure 17 on page 26.
Figure 17 PowerVC integration as Storage manager
During the execution of the use cases, SAP LaMa calls:
•Storage commands via PowerVC to retrieve information about the storage volumes that contain SAP systems.
•Storage commands via IBM PowerVC to configure and run the cloning and copying process of SAP systems.
For more information about IBM PowerVC and SAP LaMa, see the following SAP notes:
•2372251: Integration of IBM Components into SAP LaMa for Power
•2372252: Configuration of IBM PowerVC virtualization adapter for SAP LaMa
•2075173: Integration of IBM Components into SAP LVM 2.1 for IBM Power
•2372935: Configuration of IBM PowerVC Storage Adapter for SAP LaMa
IBM FlashSystem 9200 Storage System
IBM FlashSystem 9200 systems are virtualized, NVMe-based, all-flash, software-defined storage solutions designed to unlock the power of your data by combining the latest IBM Spectrum Virtualize, IBM FlashCore®, IBM Storage Insights, and IBM FlashSystem 9200 multicloud software technologies.
The IBM FlashSystem 9200 storage system, as shown in Figure 18 on page 27, runs IBM Spectrum Virtualize software and introduces remarkable new features in comparison to the predecessor models:
•NVMe drive support: NVMe is a new logical device interface standard from 2011 for accessing non-volatile storage media that is attached via a PCI Express bus.
•Lower latencies through RDMA, direct memory access from the memory of one node into that of another without involving either operating system.
•Clustering option with the older V7000 systems.
Because traditional storage systems can provide capacity for established backup environments, the new IBM FlashSystem 9200 systems are the storage foundation for building the next generation Data Protection and Reuse solutions:
•The IBM Spectrum Virtualize code stack allows you to create snapshot copies of VMs quickly and efficiently
•IBM FlashSystem 9200 Multi-Cloud solution for data reuse, protection, and efficiency
It is recommended to use fully equipped 9200 units only, with 24 NVMe Flash Core Modules.
Figure 18 IBM FlashSystem 9200 NVMe
For more information about the IBM FlashSystem, see IBM FlashSystem 9200 and 9100 Best Practices and Performance Guidelines.
For more information about which IBM storage models are certified for SAP HANA, see the Certified and Supported SAP HANA Hardware Directory.
IBM Power System H922
IBM Performance Optimization With Enhanced RISC (POWER) IBM POWER9™ servers were launched in 2018 and are based on high-performance processors, the ninth generation of its kind. POWER9 servers are one of the most powerful, secure, and scalable servers in the world. The top-of-the-line POWER9 server model is E980, which can support up to 192 cores and 64 TB of memory. POWER9 servers also are available in entry-level and midrange models H922 and H924. POWER9 servers support IBM AIX, Linux, and IBM i operating systems.
Built and optimized for SAP HANA, the IBM Power System H922 (see Figure 19) and H924 boast a large memory footprint of up to 4 TB in a dense form factor for gaining insights faster while delivering high reliability and performance. For this Blueprint setup, a H9223-22H Power System was used and the H922 featured following configuration:
•CPU: 20
•Memory: 512 GB
•Fibre Channel adapter: 2 x PCIe2 16Gb 2-port Fibre Channel adapter
•Network: PCIe3 2-port 25/10Gb NIC@ROCE SFP 28 Adapter
Figure 19 IBM Power System H922
The communication between SAP LaMa and the H922 is done via the PowerVC server. In SAP LaMa, no configuration is required for the PowerSystem.
For more information about the IBM H922, see IBM Power Systems H922 and H924Technical Overview and Introduction.
SAP Adaptive Extensions
The SAP Adaptive Extensions (SAPACEXT*.SAR) package contains the IBM storage and platform libraries for SAP LaMa. The libraries must be installed in addition to SAP Host Agent for performing various operations using SAP Landscape Management software. You install SAP Adaptive Extensions on every managed host where SAP HANA is installed.
Download the latest patch for your platform of SAPACEXT.SAR from the SAP Help Portal. (At the portal, click Support Packages and Patches → Browse Download Catalog → SAP Technology Components → SAP ADAPTIVE EXTENSIONS.
This package contains the following main components:
•libsaposprep.o
This library includes specific functions for activating and deactivating of virtual IP addresses and for handling of IPSec firewall.
•libsapacosprep_ibm.o
This library includes specific functions for mounting and unmounting NFS/GPFS file systems.
The installation of the SAP Adaptive Extensions is described in the “Basic host configuration” on page 29.
For more information about the SAP adaptive extensions, see the SAP Help Portal and SAP Note 1759181.
SAP LaMa basic setup and configuration
This section describes how to set up SAP LaMa before any landscape entities are configured. The setup can be accessed after logging in by clicking Setup in the main menu.
The earlier style Web Dynpro view is opened for the setup GUI (see Figure 20).
Figure 20 Web DynPro Main Page
Legacy GUI
SAP LaMa 3.0 SP10 offers a new and advanced Web GUI, but the Web Dynpro interface is still available because it is required for some of the landscape configuration tasks. Some of the views of this established application are marked as deprecated, but it makes sense to re-enable them for a few reasons. For example, much of documentation (SAP notes and external docs) refer to the Web Dynpro GUI, so it makes life easier if it is available.
To enable the legacy views, navigate in the Setup sheet to Settings → User Interface. Ensure that Activate Deprecated Web Dynpro Views and Display Individual Instance Operations are enabled by selecting them.
SAP LaMa Engine Settings
The Engine Settings are used to control the communication between SAP LaMa and the landscape components. It is important to adjust these settings before starting with the integration of any component so they fit your needs.
To access the SAP LaMa Engine Settings, navigate in the Setup sheet by clicking Settings → Engine. Ensure that Expert Mode is enabled for this view by selecting the Expert Mode option (see Figure 21).
Figure 21 Enable Expert Mode Setting
Security settings
You must define common or individual credentials for the SAP Host Agents of the landscape. It is acceptable practice to define a common user and password for all Host Agents, but especially in larger environments, common users might not be allowed for security reasons, or password authentication is forbidden at all. In such cases, select the Allow Configurations or Credentials Individually per Host Agent option. By using this option, you can specify individual users and passwords or an X.509 certificate per Host Agent.
In the demonstration environment, a common user name and password is being used. In the same way, the settings for SAP Instance Agents can be adjusted to meet a higher level of security requirements.
Operation settings
Adjust these settings to enable the execution and management of multiple instances on the same host. To run multiple instances (not databases) select the Allow multiple instances on the same Host option. This option is required if you plan to run a system database with a central instance on the same host, for example.
Select the Allow Multiple Database Instances on Same Host option if you plan to run multiple database instances on the same host. For more information, see SAP Note 919135.
Select the Allow MountPoint Creation option if you want to copy or clone the S/4HANA database. During the provisioning process, missing mount points are created on the target host.
Basic host configuration
This section describes how to integrate a Linux or AIX host into the SAP landscape. SAP LaMa uses an agent for communication called SAP Host Agent. This document describes the manual installation of the SAP Host Agent, which works in the same way for Linux and AIX.
The following prerequisites must be met:
•If the host runs a firewall, ensure that TCP ports 1128 and 1129 are accessible from the LaMa system.
•Obtain SAP Software Provisioning Manager (SWPM) and Host Agent for your operating system and architecture from SAP.
For more information, search the SAP Help Portal for “SAP Host Agent”.
The SAP Host Agent can be installed manually or by using the SAP Software Provisioning Manager, which is required for other SAP installations.
Because the Host Agent software is shipped as an SAP Archive, the SAP proprietary SAPCAR.EXE application must be downloaded to unpack the archive.
Complete the following steps:
1. Unpack the agent archive into a temporary directory:
hana1: cd /tmp/host_agent_install_dir
hana1: /tmp/SWPM/SAPCAR -xvf /tmp/SAPHOSTAGENT40-40-70002261.SAR -manifest SIGNATURE.SMF
It is important to specify the -manifest option because the SAP signature file that is required for validating the software is not extracted otherwise.
2. Install the software:
hana1: cd /tmp/host_agent_install_dir
hana1: ./saphostexec -install -verify
The SAP Host Agent is installed into the /usr/sap/hostctrl directory and the sapadm user is created if it does not exist.
3. Set a password for sapadm:
hana1: passwd sapadm
If the SAP LaMa Engine was set up to use a common password for all Host Agents, the sapadm password must be set.
4. Ensure that the agent was installed correctly and is working:
hana1: /usr/sap/hostctrl/exe/saphostexec -version
*************************** Component ********************
…
**********************************************************
-------------------
SAPHOSTAGENT information
-------------------
kernel release 721
kernel make variant 721_REL
compiled on Linux GNU SLES-12 ppc64le cc4.8.5 for linuxppc64le
compiled for 64 BIT
compilation mode Non-Unicode
compile time Oct 20 2018 04:32:59
For Linux, SAP offers the SAP Host Agent software alternatively as an RPM package, which makes the installation process less manual.
Installing SAP Adaptive Extension
SAP Adaptive Extensions are required by LaMa to manage SAP HANA systems via the Host Agent. The software must be downloaded from SAP as a SAR archive.
Install the downloaded archive by using the following command:
hana1: /usr/sap/hostctrl/exe/sapacosprep -a InstallAcExt -m /tmp/SAPACEXT_44-20010482.SAR
SAP LaMa basic configuration
This section covers the SAP LaMA basic configuration.
Pool creation
SAP landscapes often contain many entities, which might become confusing quickly. Therefore, SAP Landscape is organized in so-called pools, which are a logical grouping of coherent components. Pools can be grouped in a higher level called containers, but this process is out of the scope of this Blueprint because we do not work with containers in the demonstration environment.
Creating at least one pool is required because all components that are added later require a pool to which they are assigned.
Complete the following steps to create a pool:
1. Click the Configuration icon in Web Dynpro.
2. In the Pools tab, click Add Pool.
Figure 22 Adding Pools
SAP landscape infrastructure configuration
Landscape infrastructure components are required for most of the landscape management tasks.
Although hosts and SAP instances can be configured without any infrastructure components, it is good practice to define the required infrastructure before the SAP systems are configured for SAP LaMa.
The demonstration environment includes configured network components, software repositories, virtualization managers, and storage managers. The following sections discuss the purpose of these components.
Network configuration
At least one network must be defined in an SAP LaMa environment. A network definition is a IPv4 network that contains the IPv4 broadcast address and the netmask of this network (see Figure 23 on page 33).
In an advanced configuration step (not covered here), an automated DNS update can be established. For more information, see SAP Note 1572841.
Figure 23 Adding network components
Storage Adapter configuration
Storage Adapters can be used for obtaining information about the attached storage environment. SAP LaMa offers different adapter types, which are all specific to the vendor and product.
In the demonstration environment, the IBM PowerVC Storage Adapter is used. For more information about the configuration and integration of this component, see “IBM System integration to SAP LaMa” on page 23.
Virtualization Manager configuration
Virtualization Managers often are deployed to manage a data center’s virtualization environment. SAP LaMa makes use of some of those Virtualization Managers. LaMa manages the SAP relevant parts of a virtual environment through the Virtualization Manager.
The demonstration environment integrated the IBM PowerVC Virtualization manager. For more information about the configuration and integration of this component, see “IBM System integration to SAP LaMa” on page 23.
SAP software repository configuration
SAP software repositories are also defined in SAP LaMa. They are used for system updates and tasks, such as cloning or deploying to a new system. Repositories are defined as paths, where the required software components must exist. It is the responsibility of the administrator to ensure that a repository is complete.
SAP LaMa uses some predefined provisioning types for a repository (see Figure 24).
Figure 24 Software Repositories
SAP instance configuration
SAP LaMa communicates with SAP instances via agents, SAP Host Agents, and SAP Instance Agents. Host Agents are required for performing landscape management tasks that interact with a host, such as preparing and unpreparing SAP instances and databases. SAP LaMa managements tasks that interact with the instances require the SAP Instance Agents.
Other types of communication is done via the storage and virtualization adapters because not all of the required information for all possible management tasks can be gathered from hosts and instances only.
The more provisioning tasks that are not directly related to hosts and instances require those external tools. Consider that cloning an SAP instance requires setting up the storage on the target host, including items such as network and SAN configuration.
Discover Hosts
The first step of instance integration is to define hosts. Because SAP Host Agents are running on each of the hosts, we can now configure LaMa to connect to these hosts and add them as SAP hosts. SAP host means here that the host can run SAP instances or databases.
To discover hosts in LaMa, click the Configuration icon in Web Dynpro. In the Hosts tab, select Discover (see Figure 25 on page 36).
The detection source must be specified first because we want to detect hosts. Therefore, Detect Using Host and Instance Agent must be selected. Only hosts should be detected by this step; therefore, Hosts only must be found.
The resolvable names of the hosts and the network port of the Host Agent must be specified. This is usually port 1128, which establishes a plain HTTP connection. If a higher level is required, a HTTPs connection can be chosen by using port 1129.
The HTTPS connection enforces at least TLS version 1.0. For more information about enabling TLS 1.1 and 1.2. in NetWeaver, see SAP note 2417205.
Figure 25 Search for Hosts
The Authentication Type is fixed to use the default credentials as defined by the SAP LaMa Engine setup.
The hosts should now appear in the Host Overview table, and each can now be edited for setting more features.
A wizard guides through the following configuration process (see Figure 26):
1. Change or verify the configuration. The version of the SAP Adaptive Extensions also can be queried here.
Figure 26 Configuring a detected host
2. Configure more host properties (see Figure 27 on page 38).
Settings must be defined whether that host is enabled for Adaptive Extension and for system isolation.
The formerly defined network must be assigned to the host so that SAP LaMa knows how to handle the network connection to that host when it performs system provisioning.
Host-specific storage properties (such as SAN WWPNs) are also specified here. These values can be retrieved directly from host by using the Adaptive Extensions.
Figure 27 Configuring more host properties
Discover SAP Instances
SAP Instances must be discovered by SAP LaMa before they can be integrated into the Landscape. LaMa offers different discovery sources, so there are many ways to detect and integrate systems.
Systems that are at a host can be detected by using the SAP Host Agent. However, if the current location of a system is unknown, the Landscape Scanner is used, which scans the local network for systems.
Regardless from which source a system is discovered, it is accessed by using the instance agent, which is a function or piece of software that is inside the system. Instance agents work similar to host agents. They use predefined network ports and are configured for several authentication mechanisms.
For more information about configuring Instance Agents, see SAP Note 1474866.
In the test environment, we use the basic authentication feature. However, because the databases were not yet installed by using a common user name, we must specify individual credentials for each system.
In this scenario, the detection of systems via existing Host Agents is described:
Figure 28 Discover SAP instances using SAP Host Agent
2. Click Discover to start the configuration mask for running a discovery.
3. Choose the hosts on which you want to detect systems.
At the bottom of this window are the settings for the instance agent that must be specified (see Figure 29).
Figure 29 Instance Agent credential settings
Consider the following points:
•It is important to select User Name/Password (Basic) as the Authentication Type for the Instance Agent. Other types do not work in the demonstration environment because the instances were not set up for those authentication types.
•The window shows a second set of credentials to be defined, which are Common Database Administrator Settings. These settings can be ignored for now because they work only if all eligible databases were created with the same DB admin credentials.
•The DB admin user is not required for connection to the instance agent.
•The DB admin is a user defined inside the database; the instance agent user is an operating system user.
•If we do not access the database, we do not need any DB users to know. However, the DB admin is used later by SAP LaMa for several operations.
Configure SAP instances
More configuration steps might be required after an instance is detected by LaMa. Because not all information that is accessible by the instance agent is gathered automatically, check the details of each detected system or instance (see Figure 30).
Figure 30 Detected instances overview
After selecting Edit, the additional information can be retrieved and is completed automatically.
Some extra setup is required for database instances. We did not specify any database user until now, so we must do that here.
Depending on the installation settings, a common user and password or individual users must be configured. Both Administrator and Manager users are required, and those users must be specified for each database instance. This information is used in configuring a minimum of four users:
•SystemDB Administrator
•SystemDB Manager
•Instance Administrator
•Instance Manager
The database credentials are shown in Figure 31 on page 42.
Figure 31 Database credentials
When a common user was specified, all user and password combinations are the same.
SAP LaMa use cases related to storage
This section covers the SAP LaMa use cases that are related to storage.
System Copy
With SAP Landscape Manager, you can create system copies or system clones of your SAP HANA environment at storage volume level. Although a system copy requires a different unique system ID, the system ID is not changed for the clone.
System clones are often used to create isolated testing, demonstration, or training systems while system copies are used to expand existing implementations or to create QA and Test systems.
As part of the cloning or copying, the newly created system can be isolated from the rest of the landscape to avoid conflicts with the original systems.
In this document, the storage-based System Copy of an SAP HANA system is described. The following steps are being performed by SAP Landscape Management:
1. Clone SAP System.
2. Rename cloned SAP System.
3. Conduct post-copy activities.
To perform a system copy of the SAP HANA system, the following prerequisites must be met:
•Licenses, such as third-party and SAP licenses, are necessary.
•Appropriate permissions are required, as described in the following web pages of the SAP Help Portal:
•Additional prerequisites:
◦ Valid Remote Function Call user
◦ No validation errors on the host, system, or instance, which are used for the system copy
◦ No alerts show up on the instance
•For a system copy of an SAP HANA system:
◦ The installation must be performed adaptively. For more information, see this web page of the SAP Help Portal.
◦ Software Provisioning Manager
◦ SAP HANA lifecycle management tool (hdblcm)
◦ SAP Host Agent must be able to write to the operations.d subfolder
For more information about the prerequisites check, see this web page of the SAP Help Portal.
To create an end-to-end system copy of an SAP HANA System, navigate to Provisioning in the Landscape Management system.
Select the SAP HANA system that you want to create the System Copy of in the System and AS tab and then, click Copy System (see Figure 32).
Figure 32 SAP HANA Source
Specify a new System ID and a new HANA SID and select the Pool from where the target system is provisioned. Enter a new master password for the target system, which is used as System password; for example, the database users and for the operating system (see Figure 33).
Figure 33 Basic Data of Target System
Next, select the host for the target system. You can place the system on one or several hosts.
You might want to select a host that does not have any other instances running to avoid interferences (see Figure 34, Figure 35, and Figure 36 on page 45).
Figure 34 Instances where not many resources are required, all instances can be on one host
Figure 35 Application Server, central services, and database are running on individual servers for maximum performance
Figure 36 For high performing systems pick a host with 0 number of instances running on to avoid inferences
With SAP Landscape Management, an adaptive setup is recommended, which means that application servers, central instances, and databases get their own virtual host names to disconnect the applications from the physical host. It allows the use of virtual IP addresses when applications are moved from one host to another. SAP Landscape Management configures the necessary DNS entries. For systems that are not adaptively installed, it is not possible to change the host name (see Figure 37).
Figure 37 Virtual host names
Next, the instance number can be changed, which might be necessary when several QA or development systems are consolidated on a single host (see Figure 38).
Figure 38 Changing the instance number
For a storage-based system copy, the complete data is copied from one volume to another volume by using the Storage adapters. The needed volumes can be selected (see Figure 39 on page 46).
Figure 39 Storage Based System Copy
To create a consistent copy, the database must be prepared; for example, all data must be written to the disk first. This process can be done online or offline and the execution can be scheduled (see Figure 40).
Figure 40 Options to create a consistent database
Figure 41 shows an example of how to schedule the copy creation.
Figure 41 Schedule Database Copies
In the Users tab, you can change or create users for the new system. If you configured your user management system in SAP Landscape Management, SAP Landscape Management creates users for you and assigns them to the correct groups (see Figure 42).
Figure 42 Configure Users
The rename functionality is using the Software Provisioning Manager to change users and passwords and file system names for the target system (see Figure 43).
Figure 43 Rename tab
To see whether the system is ready for system copy, run the prerequisite checker, which checks all instances simultaneously.
To avoid interferences of the copied system with the environment in the SAP landscape, the target system must be shielded. Therefore, SAP Landscape Management uses the firewall of the operating system to avoid any unwanted communication. In the Isolation tab, you can select ports to open necessary communication with the new system, but block all outgoing communication (see Figure 44 on page 48).
Figure 44 Configure Isolation
To make all of the necessary changes inside the target system, a default target list is provided at the source system and is run at the target system after start in the Post Copy Automation (PCA). The task list can be modified to take care of environment-specific changes.
You also can specify more than one task list, are run sequentially in the order that they are listed. However, the default task list is run on client 000 first (see Figure 45, Figure 46, and Figure 47 on page 49).
Here, you also can change the passwords for all configured remote function call (RFC) destinations to the master password.
Figure 45 One PCA task list
Figure 46 Multiple PCA task lists
Figure 47 Individual tasks of PCA task list
On the Summary tab, all entered parameters are displayed for example host names, instance numbers, and so forth. After you check the information, the System copy can be started (see Figure 48).
Figure 48 Summary Tab
The execution of all of the System Copy tasks with all individual steps can be monitored in the Monitoring view (see Figure 49 on page 50).
Figure 49 Monitoring View
Select the task in the top window to see the progress of the individual steps in the bottom window, as shown in Figure 50 and Figure 51.
Figure 50 Individual Steps of System Copy Task
Figure 51 Successful completion of the individual Steps
After all of the steps completed successfully, the System Copy Process Status changes to Completed in the Overview window of the Monitoring tab (see Figure 52).
Figure 52 Successful Completion of the System Copy Process
System refresh
Refreshing a System copy (for example, QA or Test) with data means overwriting the existing target system with the latest data from a source system while maintaining the configuration.
For refreshing the system, the following tasks are performed by SAP Landscape Management (see Figure 53 on page 52):
•Export Configuration
•Stop
•Unprepare:
– Unregister database
– Detaches DB from host
– Unbind virtual IP addresses
– Unmounts file systems
– Unprepare IP addresses
•Clone
•Prepare
•Activate Isolation
•System-rename
•Start
•Import configuration + PCA
•Deactivate Isolation
The Refresh option can be found by using the Provisioning tab of Lama.
Figure 53 SAP Landscape Management Overview window
In the following example, T41 is a system copy of S41 and is used for QA. To update (for example) a previously copied or cloned system T41 with all the changes of S41, the refresh system functionality can be used (see Figure 54).
Figure 54 Configure System Refresh
The following options are available:
•Refresh System: Refreshes the complete system, including central instance, application server, and database.
•Refresh database: Refreshes only the database. Configuration settings of the application instance and central instance are kept.
•Restore-refresh Database: For SAP HANA databases only.
Refresh is done by restoring a backup of the source system (for example, the production system) into the target system (for example, the QA System). The configuration settings of application instance and central instance are kept.
Refresh System
During the Refresh System process, the target system is overwritten by the Source Target data.
To restore the target configuration settings after the system refresh, they must be exported first. If the configuration settings exist, they can be used by specifying the directory:
•Step 1: Basic (see Figure 55)
Figure 55 Figure 51:Basic Configuration
•Step 2: Hosts (see Figure 56)
In the next step, the source and target hosts can be selected. Because this refresh is a system refresh, SAP Landscape Management completes the current target system information but you can define a new target host that gets you a new refreshed system.
Figure 56 Host Configuration
•Step 3: Host names (see Figure 57 on page 54)
The IP addresses and virtual host names are displayed in the Hostnames tab.
Figure 57 Virtual Hostnames Configuration
•Step 4: Storage
In the Storage tab, the storage can be selected where the data is copied to and if a snapshot is sufficient or if a full copy is necessary (see Figure 58).
Figure 58 Storage Configuration
•Step 5: Consistency
The consistency options can be selected in the Consistency tab. For production environments, an online copy might be necessary; but for this, the database must be put in a consistent state (Suspend I/O Mode), which ensures that all data is written to the disk (see Figure 59).
Figure 59 Consistency Configuration
•Step 6: Users
Users can be selected in the User’s tab and passwords can be changed (see Figure 60).
Figure 60 Users Configuration
•Step 7: Rename
In the Rename tab, the system rename can be configured; for example, the directories to use (see Figure 43 on page 47). If any warnings occur that indicate that directories exist, they can be ignored because this issue is expected for a refresh.
•Step 8: Isolation
After the copy, the system must be isolated to not communicate with other systems in the landscape (see Figure 44 on page 48).
•Step 9: ABAP PCA
The ABAP Post copy automation (PCA) process makes all necessary changes in the target system; for example, imports the exported configuration settings, again via a task list (see Figure 44 on page 48 or Figure 46 on page 48).
•Step 10: Delete Storage Volumes
Configure if you want to delete the original storage volumes or to keep them.
•Step 11: Summary
After checking the system refresh configuration, the Start System Refresh can be started.
In the Monitoring window, the System Refresh Task is listed with the individual tasks in the window.
Relocate
The following describes the relocation of an instance or a set of instances from one host to another physically or virtually. Instances can be relocated in running status or not running status. The operation retains the status of the instances. Not running instances are unprepared and prepared.
Prerequisite
If you relocate SAP instances that are connected to an SAP HANA database, ensure that you installed the hdbclient with the -H <virtual-hostname> option. If this is not the case, the hdbuserstore entry is generated only for the current physical host. Check the <hdbclient-dir>/install/installation.ini file. This file must contain an HOSTNAME=<virtual-hostname> entry.
For more information, see Secure User Store (hdbuserstore) in the SAP HANA Security Guide.
Notes
SAP Landscape Management cannot relocate SAP HANA tenant databases.
An SAP HANA System can be relocated only to a target system that runs on the same hardware platform as the source system:
1. From the operations and maintenance window, select the SAP system that you want to relocate and use the drop-down arrow next to Operations to select the relocate functionality (see Figure 61).
Figure 61 Operations and Maintenance window
2. In the Setup execution window, select the target host (see Figure 62).
Figure 62 Configure Host
3. Hosts can be selected from the Allowed values window that opens. Click the icon that is next to the host (see Figure 63).
Figure 63 Available Hosts
If you select Hard Database Shutdown or Hard Instance Shutdown, the database or instance stops immediately, even if clients are connected.
Specifying Hard Instance Shutdown as “False” allows you to specify a timeout if the instance supports a soft shutdown mode.
Selecting Ignore Host Check or Ignore Intersystem Dependencies is not recommended and should be used with care.
4. Click Execute and check the progress of the individual tasks in the monitoring window.
For more information, see the User Guide SAP Landscape Management 3.0 Enterprise Edition on the SAP Help Portal.
Appendix
To ensure that SAP applications run smoothly, SAP offers the tailored data center integration (TDI) approach to hardware selection. This approach allows customers to maintain hardware flexibility with sensitivity to budget.
For a list of all storage systems certified for SAP HANA production, see the Certified and Supported SAP HANA Hardware web page (log in required).
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.