SAP HANA 2.0 Server Installation – by the SAP HANA...
Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
If you are new to PAM, see our introductory tutorial video about PAM.
SAP HANA Academy - Product Availability Matrix (PAM)
Please Note
SAP has published several Knowledge Bases Articles (KBA's) on the SAP ONE Support portal. The list below provides a starting point and is not complete.
Additionally, there are dedicated support notes for each operating system version with recommended settings, e.g. for SLES 11 SP2, SP3, etc.
A configuration guide in PDF format is attached to the notes with guidelines. These guides are not available on the SAP Help Portal for the SAP HANA platform.
Configuration Guide
Software Download
You can download the SAP HANA 2.0 platform edition from the SAP ONE Support Portal. How this works and how to distinguish the main Support Package Stack (SPS) release from subsequent revisions is the topic of the download video.
SAP HANA Academy - Download SAP HANA 2.0 SPS 02 + latest revision
Typically, the SAP HANA server will be a minimal install for security purposes (hardened) with no graphical environment and, most likely, with secured access for inbound access for administration purposes. Downloading software directly to the Linux server, for that reason, will be a very unlikely scenario.
Should you have downloaded the full SPS release to a Windows computer and struggle to get the zip files to the HANA server, you might find the next tutorial of interest. Here we show how you can upload the files using PuTTY Secure Copy (PSCP) from a Windows PC to the HANA Linux server.
SAP HANA Academy - Secure File Copy with PSCP
SAP HANA Server Installation
Once you have prepared the environment and transferred the download to the server, installing the SAP HANA server is very easy.
The tool to use is the SAP HANA Database Lifecycle Manager, 'hdblcm' on the command-line. A windows version is also available, 'hdblcmgui'. However, as most HANA systems will not have a graphical environment, you will not be using the version very often. The web version,'hdblcmweb', can be used for updates, not for an initial installation.
You might also have noticed 'hdbinst', 'hdbuninst', 'hdbupd' and 'hdbsetup'. These are the tools that the lifecycle manager calls to perform the particular platform LCM action. When performing an SAP HANA server installation, you should use hdblcm and not any of the stand-alone tools. When installing the SAP HANA client, for example, on another system, using hdbinst or hdbsetup is perfectly fine. HDBLCM is only available for server installations.
Here we are running 'hdblcm' from the HDB_SERVER_LINUC_X86-64 directory directly, Alternatively, we could have used the included Lifecycle Manager from HDB_LCM_LINUX_X86_64.
hdblcm
On the first screen, the Lifecycle Management tool will display detected components. These are all the components which you can install together with the (database) server. Should the list be empty, you can use the call option --component_root=<path> or --component_medium=<path> to specify to the installer the location of the installation media (download).
hdblcm: Choose an action
The next screen, you are prompted to provide a comma-separated list of the components you want to install. The full SPS 02 release includes the SAP HANA client, SAP HANA studio, Smart Data Access (SDA) for data virtualization, SAP HANA XS Advanced (XS Classic is included by default), SAP HANA AFL (Application Function Library) and EML (External Machine Language Libary) and, to enable access for SAP Analytics Cloud, Lumira and other EPM tools, the EPM-MDS plugin.
hdblcm: Select component
Next, you will be prompted to provide the installation path with the default mount point '/hana/shared' provided. For this, see the topic
SAP HANA is designed to scale and it is incredibly easy to create a distributed system (what other vendors might call a cluster) because the file layout is exactly the same. The program files (binaries), data files and (redo) log files are all stored in a location, a mount point, that is intended to be shared across multiple instances. Only files that are relevant to the local installation like server trace and log files, server-specific configuration files, etc. are stored on the local file system which defaults to '/usr/sap'.
After entering the installation path and, typically, accepting the value for the local hostname, you will also be prompted if you want to add more hosts. This is where you add the hana2 and hana3 below and this is the only difference between installing SAP HANA on a single server or installing on 3, 5 or 20 servers. Amazing!
Next, you are prompted to enter the SAP System ID (SID) and instance number. The SID must be three alphanumeric characters with uppercase letters and start with a letter with some restrictions: SAP or SID are off-limits. The SID must be unique throughout your organization. Choose with care, as the SID will be included in several configuration items and changing the SID will require downtime.
The instance number must be a two-digit number between 00 and 97. In a distributed system, both SID and instance number will be the same.
sid - SAP HANA Server Installation and Update Guide
Next, you are prompted to provide the worker group of the server (localhost). The default value is 'default' which will also be the most common value used here for this parameter. Alternative values are, for example, 'standby', 'extended_storage_worker' (for Dynamic Tiering), 'xs_worker' (for XS Advanced), etc.
Next, you are prompted to select the 'system usage' with 'custom' default value. Only when selecting 'Production', this parameter has an effect on the behavior of certain client tools. For example, the system will automatically be added to the Production group of SAP HANA cockpit, and, when executing SQL, a warning will be triggered. The value will be stored in a system parameter and can easily be modified after the installation: global.ini > [system_information] > usage
hdblcm: continued
Data and log volumes default to the same shared mount point location with the SID appended.
For non-productive usage, you can install multiple HANA systems on a single host (assuming enough memory is available to support such a configuration). In this case, you might want to restrict the maximum memory allocation. Another reason might be that you have licensed SAP HANA for x GB whereas the server has y GB of RAM. Again, this is governed by a system parameter: global.ini > [memorymanager] > global_allocation_limit
Next, you will be prompted for the certificate hostname for the server. This is where you specify how the HANA server will be known on the internet (or intranet), the public name. This way, when you connect with SAP Web IDE for SAP HANA to http://mydevelopdb.myorgname.com you will not get any invalid hostname certificate warning when the actual hostname is AXY5676.dummydomain.
Certificates: There is a problem
Next, you will be prompted to enter the password for the operating system administration account, <sid>adm. This is a local Linux account. You can change the password at any time using the Linux 'passwd' command. The Linux password policy defines what are valid entries and what not.
The system administrator home directory, login shell, and user ID default values are perfectly fine for 99% of the installation scenarios. The Linux user ID needs to be the same on all hosts of a distributed system and starts at 1000 for the next available ID. Assuming all hosts of a distributed system are the same, this will not be an issue but should, on one of the systems, additional users have been created, you might want to start at a higher number, e.g. 7000.
You will also be prompted to enter the password for the database superuser SYSTEM. This will be the password for both the system database and the default <SID> tenant database. Here the SAP HANA password policy applies, default: [Aa1], 1 uppercase and 1 lowercase letter and one digit.
For productive environments, SAP recommends to deactivate the SYSTEM user and to create lesser privileged users for particular purposes:
Finally, you will be prompted if you want the SAP HANA database to start when the operating system starts (the actual question is 'Restart system after machine reboot' but also covers system startup). This defaults to no, which means that you will have to start the SAP HANA database yourself after Linux has started.
Configuring this behavior afterward is not as straightforward as you might expect. You will need to edit the autostart parameter in the profile. There is no graphical user interface for this.
<sapmnt>/<SID>/profile/<SID>_HDB<instance_number>_<hostname>
When all parameters have been provided, the ominous sounding "Summary before execution" is displayed asking you if you want to proceed or not.
hdblcm: Do you want to continue (y/n)
On Your Mark, Get Set, Go!
The actual installation of SAP HANA server is a quick process. The files are extracted (preparing package ...), copied to the installation directory (installing package ...), after which the system is created and the processes started.
hdblcm: Starting processes
For a list of the different SAP HANA services and the role they perform, see
When all processes have started, default content is imported. This concerns web applications for SAP HANA XS, classic method. Some delivery units are only available for downward compatibility reasons and do not contain any content (HANA_ADMIN for SAP HANA cockpit 1.0, for example).
Finally, a resident version of the SAP HANA Database Lifecycle Manager tool is installed. This tool can be used for any subsequent platform lifecycle management tasks like adding hosts or renaming the SID.
We can also see that SAP HANA host agent configurations are deployed. Each SAP HANA server installation includes the SAP HANA host agent.
There is a feedback form to share your experiences and a log file, should any issue have occurred.
Troubleshooting - SAP HANA Server Installation and Update Guide
hdblcm: Log file written to ...
What's Up?
With the 'HDB info' command you can check the status of the different SAP HANA server processes. The 'sapstartsrv' process will have started the hdbdaemon process which, in turn, will have started the system database (hdbnameserver), the tenant database (hdbindexserver) and different supporting processes.
Over the years, for the SAP HANA Academy, SAP’s Partner Innovation Lab, and à titre personnel, I have written a little over 300 posts here for the SAP Community. Some articles only reached a few readers. Others attracted quite a few more.For your reading pleasure and convenience, here is a curated list of posts which somehow managed to pass the 10k-view mile stone and, as sign of current interest, still tickle the counters each month.