Huawei SAP HANA Appliance Technical White

Huawei SAP HANA Appliance

Technical White Paper single node Solution

Copyright © Huawei Technologies Co., Ltd. 2016. All rights reserved.

No part of this document may be reproduced or transmitted in any form or by any means without prior written consent of Huawei Technologies Co., Ltd.

Trademarks and Permissions

and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.

All other trademarks and trade names mentioned in this document are the property of their respective holders.

Notice

The purchased products, services and features are stipulated by the contract made between Huawei and the customer. All or part of the products, services and features described in this document may not be within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements, information, and recommendations in this document are provided "AS IS" without warranties, guarantees or representations of any kind, either express or implied.

The information in this document is subject to change without notice. Every effort has been made in the preparation of this document to ensure accuracy of the contents, but all statements, information, and recommendations in this document do not constitute a warranty of any kind, express or implied.

Huawei Technologies Co., Ltd.

Address: Huawei Industrial Base

Bantian, Longgang

Shenzhen 518129

People's Republic of China

Website: http://enterprise.huawei.com

Issue 01 (2013-01-2512) Huawei Proprietary and Confidential

Copyright © Huawei Technologies Co., Ltd. i

Huawei SAP HANA Appliance Technical White Paper single node edition About This Document

About This Document

Purpose

This document describes the Huawei server-based SAP High-Performance Analytic Appliance

(HANA) Scale Up solution in terms of the basic architecture, innovative high-performance inmemory database technologies, and application modes.

Intended Audience









This document is intended for the audience who want to get familiar with the Huawei

RH2288H V2 rack server-, E9000 blade server-, RH8100 V3 rack server- and RH5885H V3 server-based SAP HANA solution.

The audience are assumed to be familiar with the following products:

Huawei RH2288H V3 rack server

Huawei E9000 blade server

Huawei RH5885H V3 rack server

Huawei RH8100 V3 rack server


Copyright © Huawei Technologies Co., Ltd. ii

Huawei SAP HANA Appliance Technical White Paper single node edition Contents

Contents

About This Document.....................................................................................................................ii

1 SAP HANA..................................................................................................................................... 1

1.1 Overvie w ............................................................................................................................................................................. 1

1.2 SAP HANA Architecture .................................................................................................................................................. 1

1.2.1 SAP HANA Database ............................................................................................................................................. 2

1.2.2 SAP HANA Application ......................................................................................................................................... 3

1.2.3 SAP HANA's Me mory-based Innovative Technologies ................................................................................... 4

1.2.4 SAP HANA Application Scenarios ...................................................................................................................... 6

1.2.5 SAP HANA De livery Mode .................................................................................................................................. 8

1.2.6 SAP HANA HA Solution ....................................................................................................................................... 9

2 Huawei Solutions........................................................................................................................ 12

2.1 Huawei SAP NANA Solution ........................................................................................................................................ 12

2.1.1 Overvie w................................................................................................................................................................. 12

2.1.2 RH5885H V3 Server SAP HANA solution....................................................................................................... 12

2.1.3 RH8100 V3 Se rver SAP HANA Solution ......................................................................................................... 14

2.2 Huawei SAP NANA HA Solution................................................................................................................................. 17

2.2.1 System Replication Solution ............................................................................................................................... 17

2.2.2 Host Auto-Failover So lution................................................................................................................................ 21

2.3 RH5885H V3 .................................................................................................................................................................... 21

2.3.1 RH5885H V3 Structure ........................................................................................................................................ 22

2.3.2 RH5885H V3 Features ......................................................................................................................................... 23

2.3.3 RH5885H V3 Advantages.................................................................................................................................... 25

2.3.4 RH5885H V3 Technical Specifications ............................................................................................................. 30

2.3.5 RH5885H V3 Server Management..................................................................................................................... 33

2.4 RH8100 V3 ....................................................................................................................................................................... 35

2.4.1 RH8100 V3 St ructure ........................................................................................................................................... 35

2.4.2 RH8100 V3 Features ............................................................................................................................................. 36

2.4.3 RH8100 V3 Advantages ....................................................................................................................................... 39

2.4.4 RH8100 V3 Technical Specificat ions ................................................................................................................ 40

2.4.5 RH8100 V3 Se rver Management ........................................................................................................................ 42

2.5 ES3000 SSD ..................................................................................................................................................................... 44


Copyright © Huawei Technologies Co., Ltd. iii

Huawei SAP HANA Appliance Technical White Paper single node edition 1 SAP HANA

1

SAP HANA

1.1 Overview

Along with scientific technology development, a large volume of data is generating no matter in a personal life or enterprise operation. As data volume gets larger and larger, commercial data analysis tools keep developing rapidly. Facing the ever-changing market environment and competition and customer requirements, enterprises are in need of an efficient data analysis and processing system. This system can analyze service data and then rapidly provide the processed results while receiving data. Using the analysis results combining with other elements, such as the market conditions, customer requirement trends, and price fluctuation, decision makers can make quicker and better market and development strategies.

To fulfill customer requirements, the Huawei RH5885H V3 rack server (RH5885H V3 for short) and the Huawei RH8100 V3 rack server(RH8100 V3 for short) in the server-based SAP

HANA solution provides strong computing and expansion capabilities using the latest Intel

®

Xeon

TM

Processors E7 v3 Family. The Huawei-developed Peripheral Component Interconnect

Express (PCIE) solid-state drive (SSD) ES3000 (ES3000 for short) provides high read and write input/output operations per second (IOPS) for SAP HANA, improving SAP HANA's performance for analyzing and loading data. Customers can use SAP HANA's memory for big data analysis to accelerate business intelligence (BI), reporting, and analysis tasks. Working with the SAP BusinessObjects solution, SAP HANA supports a high-speed data warehouse and a powerful computing engine, providing a comprehensive SAP enterprise solution based on SAP HANA. In addition, SAP HANA provides an access interface for third-party software, supporting integration of different customer systems.

1.2 SAP HANA Architecture

As an application solution with integrated software and hardware, SAP HANA provides a flexible, real-time, data-source-independent big data analytic platform. Without the need for data modeling and aggregation, customers can directly query and analyze a large volume of service data. Hardware devices for SAP HANA are certified by SAP HANA, provided by

SAP's hardware partners, and commissioned together with SAP HANA to achieve the optimal performance. Running in the memory, SAP HANA database is an object-oriented database technology based on rows and columns. Using the most advanced hardware parallel processing technology, SAP HANA ensures analysis for a large volume of data and real-time query.

Figure 1-1 shows the SAP HANA architecture.


Copyright © Huawei Technologies Co., Ltd.

1

Huawei SAP HANA Appliance Technical White Paper single node edition

Figure 1-1 SAP HANA architecture

1 SAP HANA

The SAP HANA platform consists of the SAP HANA in-memory database, modeling tool,

and client tool, as shown in Figure 1-1. It processes CRUDQ operations using the HANA

computing engine and supports Structured Query Language (SQL), Multidimensional

Expressions (MDX), SAP data, and non-SAP data. Data in SAP's in-memory database is not only stored in the memory but also written to hard disks. This ensures data security.

1.2.1 SAP HANA Database

SAP HANA database is the core component of the SAP HANA platform. It comprises of the following functional components:



Row store and column store used in the same relational database engine: In the SAP

HANA database, customers can specify row store or column store for a table.

− Row store

−

Similar to the traditional databases, SAP HANA database stores data based on rows.

Row store provides high write performance and supports lower data compression rate and query performance than column store.

Row store is recommended for the following scenarios:





A small volume of data is stored in a table.

A program processes a record at a time.





A program accesses all recorded data regularly.

A column contains different values, and the data compression rate of the table is low.



Aggregation and quick search are not required.

Column store



2


Column store is used for storing tables to provide high performance for large data tables, and data aggregation, compression, and analysis.

Column store is recommended for the following scenarios:



Computing operations are executed on a single column or several columns.





The search operation is executed based on values only in several columns.

A table has many columns.





A table has many rows, and operations based on columns need to be executed, including aggregation, search, and scanning.

Most columns contain only a few different values, and the data compression rate of the table is high.







Computing engine and external interfaces: SAP HANA supports access via various interfaces. Databases such as SQL (JDBC or ODBC), MDX (ODBO), and BICS (SQL

DBC) are supported. The computing engine can directly operate in the in-memory database without the need for removing data to the application layer. SAP HANA uses the unique SQL scripts as the extended SQL statements to import high-density service data at the application logic layer to the SAP HANA in-memory database.

Data security: The SAP HANA database provides a logic layer with permanent data storage. All service data is not only stored in the memory but also synchronized to hard disks. SAP HANA database provides the log recording function. Every transaction and operation are recorded and saved to the data storage zone and log storage zone of the logic layer with permanent data storage. The log volume of the log storage zone uses the high-performance solid-state flash storage media and technology, providing high I/O performance and low latency and ensuring data security.

SAP HANA database has several management components, including data transmission management, database rights management, metadata management, upper-layer session management. Using these management functions, SAP HANA database works efficiently and securely, and provides various relational data management and rights management functions.

1.2.2 SAP HANA Application

The SAP HANA application layer has the components such as the SAP HANA database,

Software Update Manager, SAP Host Agent, SAP HANA Client, SAP HANA Studio, as

shown in Figure 1-1. These components operate and manage the SAP HANA database.

Management and installation files based on the Eclipse tool provide simple methods to enable customers' applications to connect to the SAP HANA database. Other than SAP HANA

Client, a series of library files are required for connec ting to the SAP HANA database. In general, SAP HANA Studio, SAP HANA Client, and library files are installed in the client terminal computers or servers.

SAP HANA's Software Update Manager automatically downloads and updates software in the

SAP HANA architecture. All update software can be downloaded on SAP's official website or through other channels using SAP Host Agent. SAP HANA also provides a life cycle manager for managing life cycles of software in the SAP HANA architecture. In addition, Software

Update Manager can be used to update this manager.



3


1.2.3 SAP HANA's Memory-based Innovative Technologies

Table 1-1 SAP HANA's innovative technologies

Technology Description

Each database system supports both row store and column store, making SAP HANA suitable for the system application architecture with mixed loads.

Column and row store

Minimal projections

Insert only

Bulk load

The property association technology helps automatically create and maintain the combination main key, which can combine several data columns into one column. This technology enables faster access performance for simple search.

The insert only technology implements all data update operations, and uses multiversion concurrency control

(MVCC) and asynchronous increment combination to completely solve problems of online transaction processing

(OLTP) applications that are supported by column store.

Besides business-type data insertion, HANA also supports bulk load in memory. This technology enables a large volume of data to be written to HANA without using business resources and supports super high loads for data loading.

Multi-core/parallelization

Active/passive & data aging

This technology fully utilizes advantages of the multi-core, large-memory architecture and manages processor threads at the software layer, which perfects parallelization. Each server node in the HANA cluster can access only the local memory.

Data in memory can be released based on requirements. Only the frequently used data is stored in memory. The system automatically stores data in the memory or puts the data into sleep at the persistence layer. Users can manually choose to store data in memory or hard disks.

Partitioning

In-memory compression

This technology enables data to be divided vertically and horizontally and evenly distributed to each server node in the

HANA cluster. This maximizes utilization of all system computing resources. If the data needs to be redistributed later, use data redistribution based on the in-memory computing technology, which distributes the data faster than traditional technologies.

This technology enables data to be compressed in memory, which facilitates data transmission to processors in the unit time. Then, processors can process data faster, such as scan and aggregate data.

This technology enables data to be processed in parallel. To evenly use all computing resources, users can set the number of threads for data processing on each node.

Multi-threading within



4


Technology

nodes

Description

No aggregates

This technology supports data query and analysis based on time periods. Users can specify various time periods and analyze data changes during these periods.

Analytics on historical data

SQL interface on columns

& rows

This technology supports SQL access to row store and column store and association computing between a row store table and a column store table.

The dynamic aggregation technology supports modeling based the most detailed data without using materialized views and aggregation tables, generating physical intermediate computing results, or requiring redundant storage for materialized views. This ensures data consistency for various versions.

This technology enables services to be provided by a single node, a HANA cluster with PB-level memory, or a cloud platform.

Single and multi-tenancy

Reduction of tiers/layers

This technology simplifies the software architecture, eliminates materialized views, reduces data redundancy, and implements service computing at the application layer in a high-performance database.

Dynamic extensibility

MapReduce

This technology enables a database table to be instantly expanded with new columns without reorganizing the original data. Adjustments can be dynamically made to the model on modeling without affecting the front-end data access interface.

This technology implements the MapReduce programming mode in HANA and allows application R&D personnel to define map functions for processing a large volume of data in parallel without worrying about execution efficiency and parallelization of the computing engine in HANA.

This technology uses the group key defined in memory to associate the service data index table with the detail item table, which enables the access to a detail item without scanning the entire table.

Group key



5


1.2.4 SAP HANA Application Scenarios

SAP HANA can be used in multiple application scenarios. Customers can analyze the existing

IT deployment situation and choose the SAP HANA configuration. SAP HANA can be used as a data market, enterprise-level data warehouse, traditional system with SAP application acceleration, high-performance database, and memory application platform.



SAP HANA data market

Figure 1-2 In a traditional relational database management system (RDBMS) or SAP-based

application platform for enterprise, BI is used for helping data analysis and decision making. Data used for BI analysis can be collected and processed by a data market or enterprise-level data warehouse, such as SAP NetWeaver Business WareHouse (SAP BW). Through extract, transform, and load (ETL), data of related service systems is extracted to the data market or enterprise data warehouse. Typical data market application

A large volume of data needs to be extracted from the enterprise's data warehouse, and the data results are displayed by using queries or reports. This mode requires a long response time. SAP NetWeaver business warehouse accelerator (BWA) uses memory technologies to accelerate the query and report software solution for the enterprise data warehouse. SAP NetWeaver BWA is suitable for enterprise-level data warehouse applications but may not be suitable for small-scale data market applications. As a highperformance in-memory computing platform, SAP HANA efficiently solves these problems by combining BI or upper-layer service applications. For example, the SAP

BusinessObjects tool copies data at the system layer in real time to SAP HANA. Then BI or the upper-layer application tool can provide real-time data analysis.

Figure 1-3 shows the SAP HANA data market application.



6


Figure 1-3 SAP HANA data market application

1 SAP HANA

As shown in Figure 1-3, SAP HANA supports SAP's enterprise application systems and

third-party enterprise application system. SAP provides the BI Consumer Services

(BICS) interface for the SAP BusinessObjects tool and the SAP BusinessObjects

Explorer rapid deployment based on SAP HANA. SAP HANA provide a platform for enterprise application systems, such as enterprise resource planning (ERP), customer relationship management (CRM), and supply chain management (SCM), improving the real-time analysis capabilities of original SAP enterprise application systems and accelerating the report displaying speed.

SAP HANA supports enterprise applications with the BI purpose. After data at the system layer is copied to SAP HANA, the upper-layer ERP system sends data query, categorization, and analysis requests to SAP HANA. Then the ERP system presents data analysis, content, and reports, which can be implemented by developing a new application on the ERP system or using the rapid modeling function of the SAP ERP system.

Figure 1-4 New application of SAP HANA



SAP HANA high-performance database application

As a new-generation platform for the enterprise-level data warehouse, SAP HANA not only serves as a high-performance underlying in-memory database but also supports



7

Huawei SAP HANA Appliance Technical White Paper single node edition 1 SAP HANA various application interfaces. For example, SAP NetWeaver BW is the most comprehensive and powerful data warehouse product that supports SAP Business Suite.

Figure 1-5 SAP HANA BW application



SAP HANA development platform

In SAP's ecosystem, SAP HANA is an open platform that supports third-party BI or a new enterprise application system. As an analysis or high-performance database platform, SAP HANA supports various application scenarios that are not associated with

SAP through data synchronization. The service scenarios for SAP HANA as the development platform are as follows:

− Client-based applications: These applications can access the data models, such as database tables and views, in SAP HANA through open database connectivity

(ODBC), Java database connectivity (JDBC), ODBO, SQL, or Native SQLScript.

− Web-based applications: The XS Engine in SAP HANA serves as the application server for running web applications. Clients use standard OData and XML for

Analysis (XMLA) interfaces to access data in SAP HANA, use SAP ToolKit as the web development tool, and use Server-side, JavaScript, and Native SQLscript for storage.

1.2.5 SAP HANA Delivery Mode

SAP HANA is delivered with integrated software and hardware. SAP HANA has configuration requirements for hardware platforms. All hardware vendors need to configur e hardware based on SAP HANA's hardware configuration requirements. The hardware platforms must be certified by SAP HANA to ensure the required performance of SAP

HANA. Only the certified hardware can be approved and maintained by SAP and hardware vendors.

As an important hardware partner of SAP, Huawei provides multiple servers approved by SAP

HANA, including the RH2288H V3, RH5885 V2, RH5885H V3, RH8100 V3,and E9000.

SUSE 11 and SAP HANA are pre-installed on the Huawei SAP HANA appliance.

Figure 1-6 shows the maintenance process for the Huawei SAP HANA appliance. If

customers encounter problems in SAP HANA, SAP's aftersales maintenanc e personnel locates and analyzes the problems. If the problems lie in hardware, Huawei's aftersales



8

Huawei SAP HANA Appliance Technical White Paper single node edition 1 SAP HANA maintenance personnel handle the problems. If the problems are SUSE 11 problems, SUSE

Linux's maintenance personnel handle the problems.

Figure 1-6 Maintenance process of Huawei SAP HANA appliance

1.2.6 SAP HANA HA Solution

Issues such as disaster recovery (DR), faults, and high availability (HA), service continuity, data security, and problems such as how to quickly recover the production system after a fault or disaster occurs are customers' main concerns. SAP HANA provides HA solutions of various levels to solve reliability problems including single point of failures (SPOFs) and DR.



DR key indicators

In the IT industry, customers generally use recovery time objective (RPO) and recovery point objective (RTO) as two key indicators to evaluate the DR capability of a system.

Figure 1-7 shows the RPO and RTO indicators when a system recovers from a fault.

Figure 1-7 PRO and RTO

RTO indicates the period from the time when an IT system breaks down and services stop to the time when the system recovers to support department operation.

RPO indicates the extent to which the system and application data recovers to support department operation. The recovery point can be the backup data of the last week or the real-time data of last transaction.



9


SAP HAHA supports various DR solutions, which use different RTOs and RPOs and

require different costs. Figure 1-8 shows the DR solutions. These DR solutions provide

the following features:



Backups



SAP HANA provides the data backup function. All data in the in-memory database is stored to hard disks as logs and data. When a fault occurs in the SAP HANA system, the system can recover with data backed up in a specified time or the last backup data.

However, the data loss risks exist in certain time periods for this DR solution, and the recovery time is long.

Storage Replication



The Storage Replication solution continues back up stored data and therefore requires a storage-level data backup solution from SAP HANA's hardware vendor. To ensure data security, this solution backs up data volumes or file systems to a remote DR site over network transmission using synchronization. This backup mode requires that the distance between the primary site and the secondary site does not exceed 100 km considering the delay in data synchronization and ensuring high-bandwidth network transmission.

This DR solution provides a data loss rate close to 0, requires the medium recovery time.

System Replication



The System Replication solution is a preferred HA solution for SAP HANA because it is compatible with most solutions from SAP HANA's hardware vendors. This solution supports N+N redundancy. The SAP HANA server in the standby state at the DR site maps to the SAP HANA server at the production site. When a fault or disaster occurs on the SAP HANA server at the production site, the SAP HANA server at the DR site immediately takes over services, ensuring service continuity and data security.

Synchronous data transmission and asynchronous data transmission are supported to prevent data loss. However, the service delay is caused by the data synchronization duration. Therefore, the System Replication solution requires that the distance between the production site and DR site cannot be too large so as to ensure the network transmission bandwidth between two sites. If the distance between two sites is large, asynchronous data transmission is recommended so as to ensure servic e continuity. This solution provides a data loss rate close to 0, requires the fast recovery time.

Host Auto-Failover

The Host Auto-Failover is a DR solution for the local site and supports N+M redundancy. A standby node is added to the SAP HANA cluster and configured the standby mode for the standby node. When a fault occurs on the active node, the system activates the standby node for replacing the active node. When the standby node runs in standby mode, the standby node does not accept or handle service requests. The connection between the active node and standby node after the failover is supported by the upper-layer network. This solution provides a data loss rate close to 0, requires the medium recovery time.



10


Figure 1-8 SAP HANA HA solution description

1 SAP HANA



11

Huawei SAP HANA Appliance Technical White Paper single node edition 2 Huawei Solutions

2

Huawei Solutions

2.1 Huawei SAP NANA Solution

2.1.1 Overview

Huawei and SAP is an important partnership, Huawei can provide customers with SAP

HANA all Sizing configuration solutions, including SOH, BWOH and TDI best hardware configuration, customers can select the appropriate SAP HANA Huawei according to the company business needs hardware configuration. Huawei to provide integrated solutions for

SAP HANA, including servers, storage devices and switch devices, while providing high reliability SAP HANA solution, disaster recovery solution.

2.1.2 RH5885H V3 Server SAP HANA solution

The Huawei SAP HANA solution uses the Huawei new -generation RH5885H V3 rack server and ES3000, providing high performance and high IOPS. The RH5885H V3 is a standard 4U rack server. It uses Intel

®

Xeon™ E7-8800 V3 series processors and provides 96 DIMM slots with the maximum memory capacity of 6 TB, meeting requirements on large memory capacity and high-performance computing for the SAP HANA in-memory database. The

RH5885H V3 is configured with the ES3000, providing high IOPS for the SAP HANA database logs and ensuring read and write log operations and recovery performance of the

SAP HANA database.

The Huawei RH5885H V3 server-based SAP HANA appliance provides various size options based on the configuration versions: high-performance configuration and common

configuration, as shown in Figure 2-1 and Figure 2-2. Customers can choose the size and

version based on service and configuration requirements. In addition, the RH5885H V3 server-based SAP HANA appliance can be smoothly upgraded, which maximizes the return on investment (ROI).



12


Figure 2-1 Huawei RH5885H V3 server-based SAP HANA of high performance configuration

HANA Data Volume configure 900GB HDD disk and Log Volume 800GB SSD Card




13


Figure 2-2 Huawei RH5885H V3 server-based SAP HANA of common configuration

HANA Data and Log Volume configure 900GB HDD disk


2.1.3 RH8100 V3 Server SAP HANA Solution

RH8100 V3 server is Huawei next-generation 8-way servers, using the Intel E7-8800 V3 series CPU, the modular design, can be SAP HANA Scale up is calculated by adding expansion modules, up to 12TB of memory, fully meet the SAP HANA Application large memory requirements. Huawei FusionPar RH8100 V3 uses technology that enables hard partitioning. RH8100 V3 achieve in a single system mode and dual mode system modes. SAP

HANA support Scale up program for all configuration, support 6T HANA Sizing configure , the specific configuration specifications, please refer to Figure 2-3 and Figure 2-4.



14


Figure 2-3 Huawei RH8100 V3 server-based SAP HANA of high performance configuration





15



Figure 2-4 Huawei RH8100 V3 server-based SAP HANA of common configuration




16



Huawei RH8100 V3 server supports all SAP HANA SoH or BWoH business standard configuration, and certified by SAP HANA, SAP on the official website you can check to think about configuring authentication configuration information.

The Official website: http: //scn.sap.com/docs/DOC-52522

2.2 Huawei SAP NANA HA Solution

The Huawei server-based SAP HANA appliance supports all SAP HANA HA solutions. HA solutions of various levels use different RTOs and RPOs and require different costs.

Customers can choose their HA solutions based on actual requirements.

2.2.1 System Replication Solution

The System Replication DR solution of SAP HANA supports DR in N+N redundancy mode at the same site or two different sites, and synchronous and asynchronous data transmission.



17


Figure 2-5 System Replication logical architecture

2 Huawei Solutions

After the synchronization function is established between host 2 at the primary site and host 3 at the secondary site, all data and logs on host 2 are synchronized to host 3, as shown in

Figure 2-5. When a fault occurs on SAP HANA of host 2, the SAP HANA management

platform transfers host 2 services to host 3 using scripts in synchronization mode. No data will be lost.

Figure 2-6 System Replication solution networking

The System Replication solution supports two network configuration transfer schemes:



IP address-based redirection

If a customer's primary SAP HANA system and the secondary SAP HANA system are in the same layer-2 switching network, the IP address-based redirection scheme can be



18

Huawei SAP HANA Appliance Technical White Paper single node edition 2 Huawei Solutions used. First, a floating virtual IP address is created using the domain name service (DNS), and the IP address is mapped to another MAC address on the layer-2 switch. When the primary SAP HANA system is faulty, the IP address is mapped to the MAC address of

the secondary SAP HANA system on the layer-2 switch, as shown in Figure 2-7.

Figure 2-7 SAP HANA IP address-based redirection



DNS-based redirection

If a customer uses a layer-3 switching network, the DNS-based redirection scheme can be used. Redirection is implemented by configuring the mapping between domain names and IP addresses. When the primary SAP HANA system is faulty, the DNS configuration can be modified manually or by using scripts. The domain name of the primary SAP HANA system

can be translated to the IP address of the secondary SAP HANA system, as shown in Figure 2-

8. To implement redirection by using DNS, the client link to the SAP HANA server does not

need to be reconfigured.

Figure 2-8 SAP HANA DNS-based redirection

Huawei SAP HANA System Replication Auto takeover solution



19


Huawei provide the SAP HANA System Replication Auto takeover solutions based on SUSE

HA Cluster technology, shown in Figure 2-9. SUSE HA Cluster agent monitoring program by

SAP HANA for SAP HANA database state monitoring, when SAP HANA Primary Instance fails, SUSE HA Cluster System Replication agent software will automatically complete the

System Replication takeover operation. The Secondary Instance will take over the business.

Because of The SAP HANA System Replication can run sync mode, the primary node data is full synchronizing the secondary node, and does not require data loading operations. This solution can provide a VIP interface to App Server and fast real-time.

Figure 2-9 SAP HANA Auto takeover solution

SUSE high availability clustering software SAP HANA System Replication automatic take over process shown in Figure 2-10; when the Primary node(Node1) failures, and SAP HANA

System Replication synchronization completion status, SUSE high availability cluster software will trigger the Secondary nodes(Node2) System Replication takeover operation, the

Node2 server becomes the Primary node. The Node1 server when the fault recovery, through the command to add programs Secondary node SAP HANA System Replication, SUSE high availability cluster will automatically recognize each server node, and performs management.

Figure 2-10 SAP HANA Auto takeover process



20


2.2.2 Host Auto-Failover Solution

The Host Auto-Failover is a DR solution for the local site and supports N+M redundancy.

After a standby node is configured in the SAP HANA cluster, if an active node is faulty, the

SAP HANA system automatically activates the standby node and switches it to the active node. All log data and database data are stored in the shared storage. No data is lost during the switchover between the standby and active nodes. The standby node in the SAP HANA system does not accept or handle service requests, and processes services after state switchover.

Figure 2-11 Failover process of the Host Auto-Failover solution

Figure 2-12 Host Auto-Failover solution networking

The SAP HANA cluster connects to the shared storage over the storage network, as shown in

Figure 2-12. Customers can deploy a network based on actual conditions. SAP HANA has

requirements on IOPS. 10GE or FC storage network is recommended for customers to ensure storage I/O operations for SAP HANA.

2.3 RH5885H V3

Building on Huawei's rich experience in servers, the RH5885H V3 is a high-performance enterprise-level server that uses the latest Intel processors. TheRH5885H V3 is a standard 4U



21


4-socket rack server and supports new-generation Intel

®

Xeon

TM

E7-8800 V3 series processors. It provides 60 computing cores and a maximum memory capacity of 6 TB.

Oriented at mission-critical applications, the RH5885H V3 applies to databases, ERP, BI analysis, and virtualization.

Figure 2-13 RH5885H V3 front view

2.3.1 RH5885H V3 Structure

This section describes the RH5585H V3 in terms of its components, mainboard layout, and mainboard connectors.

Figure 2-14 RH5885H V3 structure

Issue 01 (2013-01-25)

1

3

5

7

9

Chassis

PCIe card

Standard PCIe riser card iBBU tray

(Optional) Supercapacitor

2

4

6

8

10

Huawei Proprietary and Confidential


PCIe riser card fixing frame

USB flash drive

PCIe riser card tray

(Optional) iBBU

TPM

22


11

21

23

25

27

29

13

15

17

19

Redundant array of independent disks

(RAID) controller card

Processor

Chassis cover support

M emory riser

Hard disk backplane

Hard disk

Front bezel

M emory riser fixing frame

Power supply unit (PSU)

(Optional) Secure digital SD card

2.3.2 RH5885H V3 Features

RAS Features

12

22

24

26

28

30

14

16

18

20

M ainboard

Heat sink

M emory riser tray

Dual in-line memory module (DIM M )

DVD-ROM drive

Fan module

Customized label

Network interface card (NIC)

Hot-swappable PCIe riser card

(Optional) SATA DOM

The RH5885H V3 provides the following RAS features to ensure stable system operation, simplify serviceability, and prolong the system operation time:























The eMCA mechanism automatically rectifies correctable errors to ensure normal system operation. For uncorrectable errors, you can isolate or replace the faulty component online, and configure the new component without a system restart. BIOS preferentially deals with correctable memory errors and locates the faulty DIMM.

The RH5885H V3 provides chip-level fault tolerance (such as automatic recovery from processor, chip, and link hardware faults), minimizing system breakdown caused by hardware faults.

The RH5885H V3 supports single device data correction (SDDC) and double device data correction (DDDC) to rectify memory soft errors.

The RH5885H V3 provides memory mirroring and memory sparing functions to eliminate system downtime caused by uncorrectable memory hardware errors.

The RH5885H V3 supports faulty DIMM indication on an offline memory riser, which allows faulty DIMMs to be identified on a removed memory riser.

The RH5885H V3 supports full redundancy and hot-swap maintenance without opening the chassis cover for key components, such as PSUs, fan modules, and hard disks. These features enable quick replacement of faulty components without interrupting normal system operation.

The RH5885H V3 supports hot swap of some PCIe cards without opening the chassis cover, which implements PCIe upgrades and replacement without interrupting system operating.

The RH5885H V3 supports automatic disconnection from a faulty I/O device. When a fatal I/O device fault is detected, the system enters virus mode and disconnects the link to the faulty I/O device to prevent other devices from being affected.

The RH5885H V3 supports hot-swappable drives to protect data and prolong normal system running time using RAID.

The Huawei iMana software monitors system operating, triggers alarms, and performs recovery actions. This helps minimize system downtime.

Inband and out-of-band fault management software implements PFA and fault management. The software traces components, sends a precaution before a system breakdown caused by a faulty component, runs self-diagnosis, self-correction, selfrecovery, and provides maintenance tips about faulty components for maintenance



23

Huawei SAP HANA Appliance Technical White Paper single node edition 2 Huawei Solutions personnel, including offline and online operations and component replacement. PFA can be performed on components, such as processors, DIMMs, fan modules, PSUs, and hard disks.



The RH5885H V3 provides a touched diagnostic LCD diagnosis panel to facilitate fault location, which greatly shortens the system recovery time.





The optimized heat dissipation system supports long-term stable operation at 40°C

(104°F) when no GPU is configured and 35°C (95°F) when a GPU is configured.

The advanced fault tolerance, fault recovery, and key component redundancy enable system availability of 99.999%.



Huawei provides a three-year warranty for parts replacement and onsite limited repair for the RH5885H V3 used in China. Huawei provides 9 x 5 (a 9-hour-a-day, 5-day-a-week) next business day (NBD) support. Optional service upgrades are available.

Performance and Scalability

The RH5885H V3 supports the following features to ensure high performance and scalability while reducing the total cost of ownership (TCO):



















An Intel

®

Xeon

®

E7 v2 processor used by the RH5885H V3 supports a maximum of 15 cores, 37.5 MB L3 cache, and three QPI links. An Intel

®

Xeon

®

E7 v3 processor supports a maximum of 18 cores, 45 MB L3 cache, and three QPI links. These features provide outstanding system performance.

The RH5885H V3 supports a maximum of four processors and 60 cores (E7 v2 processors) or 72 cores (E7 v3 processors), which maximizes concurrent execution of multithreaded applications.

Intel

®

Turbo Boost Technology allows processor cores to run faster than the Thermal

Design Power (TDP) configuration specified frequency if the processor cores are operating below power, current, and temperature specification limits.

Intel Hyper-Threading Technology enables each processor core to run up to two threads, improving parallel computation capability.

The hardware-assisted Intel

®

Virtualization Technology (Intel

®

VT) allows operating system (OS) vendors to better use hardware to address virtualization workloads.

The RH5885H V3 can be configured with a maximum of 96 DDR3 or DDR4 DIMMs, providing 3 TB memory capacity (with 32 GB DIMMs) and thirty-two 1866 MHz memory channels. These features offer high memory bandwidth, addressing in-memory computing.

The RH5885H V3 provides 16 standard PCIe slots for installing four dual-slot, fullheight, full-length GPUs. The RH5885H V3 provides one plug-in NIC without occupying a standard PCIe slot, allowing flexible configurations of four GE ports or two

10GE ports.

The RH5885H V3 supports four dual-slot GPUs, meeting demands for graphics processing, high-performance computing, and virtual desktop.

The RH5885H V3 supports PCIe 3.0, which increases the maximum I/O bandwidth by

60% (8 GT/s per link) compared with PCIe 2.0.

Manageability and Security

The RH5885H V3 provides the following features to simplify local and remote server management:



24




2 Huawei Solutions

The RH5885H V3 supports Intelligent Platform Management Interface (IPMI) 2.0. The built-in iMana monitors server operating status and implements remote management.



An integrated industry-standard Unified Extensible Firmware Interface (UEFI) increases setup, configuration, and update efficiency, and simplifies fault handling.





The Intel Advanced Encryption Standard New Instructions (AES NI) implement faster and stronger encryption.

The Intel Execute Disable Bit (EDB) function works with the supported OS to prevent certain types of malicious buffer overflow attacks.



The Intel Trusted Execution technology provides enhanced security by using hardwarebased defense against malicious software attacks, allowing an application to run in an isolated space from all other applications running on the OS.

Energy Efficiency







The RH5885H V3 provides the following features to reduce energy consumption and operating expense (OPEX) and increase energy efficiency:









The latest Intel

®

Xeon

®

E7-4800 or E7-8800 processors provide better performance than the previous-generation processors while fitting into the same TDP limits.

The Intel Intelligent Power Capability allows a processor to be powered on or off based on site requirements to reduce power consumption.

Low-voltage Intel

®

Xeon

®

processors consume less energy and apply to data centers and telecommunication environments that have power and thermal limitations.

The latest E7 v3 processors support 1.2 V DDR4 DIMMs, which consume approximately 20% less power with the same memory capacity and improve the operating frequency of the memory bus by around 15%.

Solid state drives (SSDs) consume 80% less power than traditional 2.5-inch HDDs.

AC and 240 V high-voltage DC power supplies are supported.

Platinum PSUs are used, achieving power conversion efficiency of 94%.

2.3.3 RH5885H V3 Advantages

The RH5885H V3 provides the following advantages:

High RAS to Improve Stability and Enable Quick Recovery













The RH5885H V3 transplants advanced RAS features from Intel Itanium (midrange computers) and implements up to 53 hardware RAS features. For details, see 1.7 RAS.

The fan modules of the RH5885H V3 are hot-swappable and can be maintained without opening the chassis cover.

Some PCIe devices used by the RH5885H V3 are hot-swappable and can be maintained without opening the chassis cover.

The RH5885H V3 comes with a touched diagnostic LCD, which offers quicker fault location and better user experience than the traditional LED panels.

The RH5885H V3 supports long-term stable operating at 40°C, while most servers support operating only at 35°C.

SDM provides the black box and KVM video recording functions, facilitating fault location for OS breakdowns.



25


Table 2-1 RH5885H V3 RAS features

No. Module Feature Name

1

Description

Processor Corrected Machine

Check Interrupt

(CMCI)

This feature corrects error-triggered interrupts.

2 DIMM Failed DIMM

Isolation

3

4

5

6

Memory Thermal

Throttling

Rank Sparing

Memory Address

Parity Protection

Memory Demand and Patrol

Scrubbing

This feature identifies the faulty DIMM, which helps isolate the faulty DIMM from others and replace it.

This feature automatically adjusts DIMM temperature to avoid DIMM damage due to overheat.

This feature uses some memory ranks as backup ranks to prevent the system from crashing due to uncorrectable errors.

This feature detects memory command and address errors.

This feature provides the memory patrol function for promptly correcting correctable errors upon detection. If these errors are not corrected promptly, uncorrectable errors may occur.

7

8

9

10

Memory Mirroring This feature improves system reliability.

Memory Board

Hot swap

This feature enables hot-swappable memory risers to be supported by the HSW-EX platform.

Intel SMI lane failover

Intel SMI Packet

Retry

This feature provides an SMI2 interface selfrecovery capability to improve system availability.

This feature provides an SMI2 interface retry mechanism upon errors to improve system reliability and availability.

11

12

13

14

SDDC

DDDC

SDDC+1

DDDC+1

This feature provides a single-device, multibit error correction capability to improve memory reliability.

This feature enables the memory to have the

SDDC function after one faulty device is replaced with a backup device to improve memory reliability and the self-recovery capability.

This feature enables the memory to correct 1bit errors after one SDDC to improve memory reliability and availability.

This feature enables the memory to correct 1bit errors after one DDDC to improve



26


No.

15

16

17

18

Module

PCIe

Feature Name

Device Tagging

2 Huawei Solutions

Description

memory reliability and availability.

This feature degrades and rectifies DIMM device faults to improve DIMM availability.

Data Scrambling This feature optimizes data stream distribution and reduces the error possibility to improve the reliability of data streams in the memory and the capability to detect address errors.

PCIe Advanced

Error Reporting

This feature improves server serviceability.

Live Error

Recovery (LER)

This feature provides a PCIe device selfrecovery capability to improve server reliability.

19

20

21

22

23

24

QPI

PCI Express Hot

Plug

Intel QPI Clock

Fail Over

Intel QPI Dynamic link retraining and recovery on link failure

Intel QPI Link

Level Retry

This feature improves QPI link reliability.

This feature provides a retry mechanism upon errors to improve QPI reliability.

Intel QPI Selfhealing

This feature improves PCIe device flexibility and server usability.

This feature provides a QuickPath

Interconnect (QPI) clock link self-recovery mechanism to improve system reliability.

Intel QPI Protocol

Protection via

CRC

This feature provides a QPI data link selfrecovery mechanism to improve system reliability.

This feature provides cyclic redundancy check (CRC) protection for QPI packets to improve system reliability.

25

26

27

28

System

Intel QPI Viral

Mode

Core Disable For

Fault Resilient

Boot (FRB)

Corrupt Data

Containment

Mode

Socket disable for

FRB

This feature provides a QPI viral mode to prevent the spreading of error data and improve system security.

This feature isolates the faulty processor during startup to improve system reliability and availability.

This feature identifies the memory storage unit that contains corrupted data to minimize the impact on the running programs and improve system reliability.

This feature isolates the faulty socket during startup to improve system reliability.



27


No.

29

30

31

32

33

34

35

36

37

38

39

40

41

Module Feature Name

Architected Error

Records

Error Injection

Support

MCA

2 Huawei Solutions

Description

With the eMCA feature, the BIOS collects error information recorded in hardware registers in compliance with UEFI specifications, sends the error information to the OS over the APEI of the Advanced

Configuration and Power Interface (ACPI), and locates the error unit, improving system availability.

This feature injects errors to verify various

RAS features.

This feature provides software recovery for uncorrectable errors, which improves system availability.

This feature improves system availability. eMCA:Gen1

MCA recovery-IO This feature integrates input/output (I/O) error reports into the MCA to allow users to process I/O device errors in a unified manner and improve system serviceability.

OOB access to

MCA registers

The out-of-band system accesses MCA registers by using the Platform Environment

Control Interface (PECI). If a fatal error occurs in the system, the out-of-band system collects onsite data to facilitate fault analysis and locating and improve system serviceability.

BIOS Abstraction

Layer for Error

Handling

The BIOS processes errors and reports the error information to the OS and iMana in compliance with specifications to improve system serviceability.

BIOS-based

Predictive Failure

Analysis (PFA)

Touched

Diagnostic LCD

The BIOS provides physical unit information for DIMM errors, and the OS traces and predicts errors, and isolates error memory pages.

The touched diagnostic LCD simplifies onsite maintenance operations and improves onsite maintenance efficiency.

Redundant Fans

Hot-swap Fans

The fan modules work in N+1 redundancy mode.

The fan modules are hot-swappable and support maintenance without the need for opening the chassis cover.

Redundant PSUs The PSUs work in 1+1 redundancy mode.

Hot-swap PSUs The PSUs are hot-swappable.



28


No.

42

43

44

45

46

47

48

49

50

51

52

53

Module Feature Name

Failed DIMM

Identification

2 Huawei Solutions

Description

Memory risers no matter they are in online or offline state support failed DIMM identification by using indicators.

This feature supports faulty fan identification. Failed FAN

Identification

Failed HDD

Identification

Failed PSU

Identification

HDD Hot Swap

Memory PFA

HDD PFA

40°C Ambiance

Temperature

This feature enables faulty hard disk identification.

This feature enables enable faulty PSU identification.

The hard disks are hot-swappable.

This feature enables precaution for memory faults.

This feature enables precaution for hard disk faults.

The RH5885H V3 supports long-term stable operation at 40°C (104°F) when no GPGPU is configured. When a GPGPU is configured, the server supports long-term stable operation at 35°C (95°F).

Software Black Box

KVM Recording SDM provides the KVM video recording function, facilitating cause analysis for system breakdowns.

The Last Screen

Server Device Management (SDM) provides the black box function, facilitating cause analysis for OS breakdowns.

SDM provides the last screen function, facilitating cause analysis for system breakdowns.

Backup image for iMana

The integrated management (iMana) software supports dual-image backup. If one image fails, the other image can be started to ensure normal software operation.

Leading Computing Performance





The RH5885H V3 uses the latest Intel

®

Xeon

®

Brickland-EX

®

(E7 v2 and E7 v3) processors with high performance. Compared with a Westmere-EX (E7 v1) processor, which provides a maximum of 10 cores and 30 MB L3 cache, a Brickland-EX processor provides a maximum of 18 cores and 45 MB L3 cache.

A Brickland-EX processor supports a maximum of 96 DIMMs. Both the number of

DIMMs and the memory capacity increase by 50% compared with the Westmere-EX processor, enabling the RH5885H V3 to support large databases and more VMs.



29




2 Huawei Solutions

Compared with a Westmere-EX processor, a Brickland-EX processor increases the overall performance by 200% and offers 340% higher performance for certain applications.

2.3.4 RH5885H V3 Technical Specifications

Table 2-2 describes the RH5885H V3 technical specifications.

Table 2-2 RH5885H V3 technical specifications

Item Specifications

Form factor/height

Processor

Chipset

4U rack server, supporting hold rails and cable management assemblies





A maximum of four Intel

®

Xeon

®

E7-4800 or E7-8800 v2

(Ivy Bridge-EX) processors, with up to 15 cores and 37.5 MB

L3 cache per processor

A maximum of four Intel

®

Xeon

®

E7-4800 or E7-8800 v3

(Haswell-EX) processors, with up to 18 cores and 45 MB L3 cache per processor

Intel Patsburg PCH

Number of DIMMs 96 slots for installing DDR3 or DDR4 DIMMs

There are 8 memory risers in total. Each processor supports two memory risers, and each memory riser supports 8 or 12 DIMMs.

Maximum memory capacity

3 TB (32 GB DIMMs)

Number of hard disks Eight or twenty-three 2.5-inch hot-swappable SAS/SATA HDDs or SSDs

Maximum local storage capacity



Eight hard disks: 8 x 1.2 TB = 9.6 TB



23 hard disks: 23 x 1.2 TB = 27.6 TB

NOTE

The RH5885H V3 supports eight or twenty -three hard disks.

RAID support A RAID controller card supports either of the following:





RAID 0, 1, 10, and 1E

RAID 0, 1, 10, 5, 50, 6, and 60, a maximum cache capacity of

2 GB, and a supercapacitor for power-off protection

Network port

Expansion slot

The onboard NIC can be flexibly configured to provide either of the following ports:





Four integrated GE 1000BASE-T ports

Two 10GE optical ports



Two GE BASE-T ports

Various PCIe risers can be used to provide several PCIe specifications:

16 standard PCIe slots



Four PCIe 3.0 x8 slots for installing full-height full-length



30


Item

2 Huawei Solutions

Specifications





 cards, supporting hot swap without opening the chassis cover

Eight PCIe 3.0 x8 slots for installing full-height full-length cards

Two PCIe 3.0 x4 slots for installing full-height full-length cards



 Four PCIe 3.0 x8 slots for installing full-height full-length cards, supporting hot swap without opening the chassis cover







Four PCIe 3.0 x16 slots for installing full-height full-length cards




 Four PCIe 3.0 x8 slots for installing full-height full-length cards, supporting hot swap without opening the chassis cover



Four PCIe 3.0 x8 slots for installing full-height full-length cards










External port





Front panel: two USB 2.0 ports, one power button, one UID button, one video graphics array (VGA) port, and one touched diagnostic LCD

Rear panel: two USB 2.0 ports, one VGA port, one serial port, one 100M management port, one UID indicator, and one plug-in NIC

DVD-ROM drive

PSU

One

The PSUs can be configured as follows:





Two 3000 W AC PSUs in 1+1 redundancy mode

Two 2000 W AC PSUs in 1+1 redundancy mode





Two 2500 W DC PSUs in 1+1 redundancy mode

Four 800 W AC PSUs in 2+2 redundancy mode

System management IPMI 2.0

Security feature Power-on password and administrator password

Front bezel



31


Item

Video card

Operating systems supported

2 Huawei Solutions

Specifications

The system mainboard integrates the display chip and provides

16 MB display memory. The maximum resolution is 1024 x 768.

Check the latest compatibility list.

Warranty

Dimensions (H x W x

D)

175 mm x 447 mm x 790 mm (6.89 in. x 17.60 in. x 31.10 in.)

Maximum weight





Packaging materials: 7 kg (15.44 lb)

Server with 23 disks (excluding the package): 56 kg (123.48 lb)

 Server with 8 disks (excluding the package): 52 kg (114.66 lb)

NOTE

When four CPUs are installed and DIM M s and hard disks are fully configured, the server is highest in weight.

The weight of PCIe cards is not included and can be calculated as follows:



HHHL PCIe card: 0.3 kg (0.66 lb)



FHHL PCIe card: 0.4 kg (0.88 lb)



FHFL dual-slot PCIe card (GPU): 1 kg (2.205 lb).

Physical environment



Operating environment

 Ambient temperature: 5°C to 40°C (41°F to 104°F)

NOTE

The maximum operating temperature is 35°C (95°F) if an PCIe SSD or a

GPU is configured.

 Ambient humidity: 10% RH to 90% RH (twmax = 29°C)





Storage environment

 Ambient temperature: –40°C to 65°C (–40°F to

+149°F)

 Ambient humidity: 5% RH to 95% RH (twmax = 38°C)

Altitude: ≤ 3000 m (9842.4 ft)

NOTE

When the altitude is higher than 900 m (2952.72 ft), the operating temperature decreases by 1°C (1.8°F) per 300 m (984.24 ft).

Acoustic noise

Three-year customer replaceable unit and onsite limited warranty,

9 x 5 NBD, and optional service upgrades

The data listed in the following is the declared A-weighted sound power levels (LWAd) and declared average bystander position Aweighted sound pressure levels (LpAm) when the server is operating in a 23°C (73.4°F) ambient environment. Noise emissions are measured in accordance with ISO 7999 (ECMA

74) and declared in accordance with ISO 9296 (ECMA 109).



Idle:

 LWAd: 7.1 Bels

 LpAm: 58.8 dBA



32


Item

2 Huawei Solutions

Specifications



Operating:

 LWAd: 7.1 Bels

 LpAm: 58.8 dBA

NOTE

The actual sound levels generated during server operating vary depending on the server configuration, load, and ambient temperature.

2.3.5 RH5885H V3 Server Management













The RH5885H V3 uses Huawei proprietary iMana 200 integrated management system to implement remote server management. iMana 200 complies with IPMI 2.0 specifications and provides reliable hardware monitoring and management. iMana 200 seamlessly communicates with management modules in a chassis and manages the compute nodes in the chassis through the management modules. iMana 200 supports the following features:

Keyboard, video, and mouse (KVM) and text console redirection

Remote virtual media

IPMI V2.0

Simple Network Management Protocol Version 3 (SNMPv3)

Common information model (CIM)

Web-based logins

Table 2-3 错误!未指定书签。 describes iMana 200 specifications.

Table 2-3 iMana 200 specifications

Item

Management interface

Fault detection

System watchdog

Boot device configuration

Specifications

iMana 200 supports a variety of management interfaces to implement system integration. iMana 200 can integrate with any standard management system over the following interfaces:





IPMI V2.0

CLI





HTTPS

SNMPv3

Detects faults and accurately locates hardware faults.

Supports BIOS power on self-test (POST),

OS watchdog, and automatic system reset for timeout. You can enable or disable these functions on iMana 200.

Supports out-of-band configuration for boot



33


Issue 01 (2013-01-25)

Item

Alarm management

Integrated KVM

Integrated virtual media

Web-based user interface (UI)

Fault reproduction

Screenshots and videos

DNS/LDAP

Dual-image backup

Asset management

Intelligent power management

IPv6

Network Controller Sideband Interface

(NC-SI)

2 Huawei Solutions

Specifications

devices.

Supports alarm management and reports alarms in various ways, such as the SNMP trap, Simple Mail Transfer Protocol

(SMTP), and syslog service, to ensure uninterrupted system operation.

Provides remote maintenance measures, such as KVM and KVM over IP, for troubleshooting. Support a maximum resolution of 1280 x 1024.

Virtualizes local media devices or images to the media devices for remote compute nodes, which simplifies OS installation. The virtual DVD-ROM drive supports a transmission rate of up to 8 MB/s.

Provides a visual WebUI for quick configuration and information queries.

Supports the following web browsers:



Internet Explorer 8.0

 Firefox 9.0





Chrome 13.0

Safari 5.1

Reproduces faults to facilitate fault diagnosis.

Allows you to view screenshots and videos without login, which facilitates preventive maintenance inspection (PMI).

Supports domain management and directory services, which significantly simplifies network configuration and management.

Starts software from an image backup if the software fails.

Provides intelligent asset management.

Uses the power capping technology to increase deployment density and the dynamic energy saving technology to lower the OPEX.

Supports IPv6 to ensure sufficient IP addresses.

Supports NC-SI, which allows you to access iMana 200 over a service network port.

Huawei Proprietary and Confidential


34


2.4 RH8100 V3

Huawei RH8100 V3 is a high-end 8-socket rack server that adopts the latest Intel

®

Xeon

®

E7-

8800 v2 processors, and uses the latest Huawei server technologies. Among the products of the same type, the RH8100 V3 has the highest reliability, outstanding performance, leading architecture, and open and easy-to-use management platform, which can meet service load requirements of key services, memory databases, virtualization, and HPC.

Figure 2-15 RH8100 V3 Server view

2.4.1 RH8100 V3 Structure

This topic describes the RH8100 V3 components.

Figure 2-17 and Figure 2-18 show the components of the RH8100 V3 with 12 hard disks.

Figure 2-16 RH8100 V3 components (front view)



35


1

3

5

7

9

11

13

Chassis

PCIe card

Supercapacitor (optional)

RAID controller card

2.5-inch hard disk

CM

Processor

Figure 2-17 RH8100 V3 components (rear view)

2

4

6

8

10

12

14

2 Huawei Solutions

DVD-ROM drive+LCD iBBU (optional)

PCIe riser module

Front I/O module of model A

Memory board

Processor heat sink

DIMM

5

7

9

1

3

Fan module 2

High-performance fusion console 4

TPM

Hot-swappable PCIe card

Rear I/O module

6

8

PSU

LOM

USB flash drive

Non-hot-swappable PCIe card

2.4.2 RH8100 V3 Features

Huawei FusionPar Technology

The RH8100 V3 employs the Huawei FusionPar technology to implement the hard partitioning function. The RH8100 V3 can operate in single-system or dual-system mode.

FusionPar changes the interconnection topology between processors based on the collaboration between the iBMC and BIOS. In single-system mode, eight processors are in the same QPI topology. In dual-system mode , eight processors are in two groups, and each group is in a QPI topology. That is, the processors interconnected through QPIs are in one partition.



36




2 Huawei Solutions

The RH8100 V3 operates in single-system mode by default. The RH8100 V3 in this mode operates as a single system.



The dual-system mode is also called dual-partition mode. The RH8100 V3 in this mode operates as two 4-socket systems: systems A and B, which are also called partitions A and B.

The hard partitioning function improves service loading flexibility. You can configure the operating mode of the RH8100 V3 based on service requirements, which helps maximize return on investment (ROI).

The FusionPar technology has the following features:













Easy switching: You can implement one-click switching between the system operating modes using the LCD or iBMC WebUI, without altering hardware components or upgrading any software.

Unified UI: When the RH8100 V3 operates in single-system mode, iBMC provides a unified UI for management.

Symmetrical distribution of service resources: When the RH8100 V3 is operating in dual-system mode, service resources are symmetrically distributed on the two partitions.

Internal integrated GE ports: When the RH8100 V3 is operating in dual-system mode, you can use the internal integrated GE ports for communication between the two partitions.

Joint control of platform resources: When the RH8100 V3 is operating in dual-system mode, fan modules and PSUs can be shared between and jointly controlled by the two partitions to achieve optimal overall performance.

DVD-ROM drive and LCD sharing: When the RH8100 V3 is operating in dual-system mode, the two partitions share the DVD-ROM drive and LCD, which facilitates local maintenance and one-click switching between the system operating modes using the

LCD or iBMC WebUI.

Performance and Scalability

The RH8100 V3 supports the following features to improve performance and scalability while reducing the total cost of ownership (TCO):











Supports a maximum of eight processors (four processors can also be configured.)

− Each Intel

®

Xeon

®

E7-8800 v2 processor ensures high processing performance by providing up to 15 cores, one L3 cache of 37.5 MB, and three 8.0 GT/s QPI links between processors.

− Each Intel

®

Xeon

®

E7-8800 v3 processor ensures high processing performance by providing up to 18 cores, one L3 cache of 45 MB, and three 9.6 GT/s QPI links between processors.

Uses Intel Turbo Boost Technology, which allows processor cores to run faster than the frequency specified in the TDP configuration if they are operating below power, current, and temperature specification limits.

Uses Intel Hyper-Threading Technology, which enables each processor core to run up to two threads, improving parallel computation performance.

Uses the hardware-assisted Intel

®

Virtualization Technology (Intel

®

VT), which allows

OS vendors to better use hardware to address virtualization workloads.

Large memory capacity and high memory bandwidth

− Supports a maximum of 192 DDR3 DIMMs with up to 1600 MT/s bandwidth when the server is configured with E7 v2 processors.



37


−

2 Huawei Solutions

Supports a maximum of 192 DDR4 DIMMs with up to 1866 MT/s bandwidth when the server is configured with E7 v3 processors.



Provides 18 PCIe slots.

−

Ten PCIe slots are on the back I/O module (BM), used for installing standard cards that need to connect to external cables, for example, NICs, HBAs, and IB cards.

− Six PCIe slots are on FM-A, used for installing standard PCIe SSD cards (Huawei

ES3000) or GPUs that do not need to connect to external cables. Up to two compatible GPUs or six standard PCIe SSD cards are supported.



− Two PCIe slots are dedicated for RAID controller cards.

Provides two LOMs with GE or 10GE ports. The two LOMs do not occupy standard

PCIe slots.

Supports PCIe 3.0, which increases the maximum theoretical I/O bandwidth by 60% compared with PCIe 2.0.

Availability and Serviceability

The RH8100 V3 provides 60 RAS features to improve availability and serviceability:

 Supports memory migration, combined memory board hot swap, to implement DIMM replacement without shutdown, which is leading in the industry.

 Complies with component selection criteria for carrier-grade component reliability and derating, and meets carrier-grade reliability test requirements.

 Implements redundancy for key components, such as the BIOS flash memory, iBMC image, PSUs, and fan modules.

 Supports faulty DIMM indication on an offline memory board, which allows faulty

DIMMs to be indicated on a removed memory board.

 Supports redundancy and hot swap without opening the chassis cover for key components such as PSUs, fan modules, and hard disks. These features enable quick replacement of faulty components without interrupting system operation.

 Supports hot swap of PCIe cards in some PCIe slots without opening the chassis cover, which implements PCIe upgrades and replacement without interrupting system operation.

 Supports hot swap of hard disks and uses RAID to protect data on hard disks and maximize system uptime.

 Uses the Huawei iBMC to monitor system parameters in real time, trigger alarms, and perform recovery actions in case of failures, minimizing system downtime.

 Integrates fault management and predictive failure analysis (PFA) to greatly improve fault locating efficiency and reduce unexpected downtime.

 Provides a 3.5-inch touch-screen LCD diagnosis panel to facilitate fault locating and shorten the fault recovery time.

 Provides a three-year warranty for parts replacement and onsite limited repair for the server, and provides 9 x 5 (a 9-hour-a-day, 5-day-a-week) next business day (NBD) support. Optional service upgrades are available.

Manageability and Security

The RH8100 V3 provides the following features to simplify local and remote server management:

 Provides iBMC, which monitors server operating status and implements remote management in compliance with Intelligent Platform Management Interface (IPMI) 2.0.



38


 Integrates the Unified Extensible Firmware Interface (UEFI)to improve setup, configuration, and update efficiency and simplify fault handling.

 Supports the Intel

®

Advanced Encryption Standard New Instructions (AES NI), which allows faster and stronger encryption.

 Provides the Intel Execute Disable Bit (EDB) function, which works with the supported

OS to prevent certain types of malicious buffer overflow attacks.

 Adopts strict security test criteria and the chip-level Huawei iBMC, ensuring secure system operating.

Energy Efficiency

The RH8100 V3 provides the following features to reduce energy consumption and operating expense (OPEX) and increase energy efficiency:

 Uses E7-8800 v2 series processors to deliver nearly double performance with only 20% higher TDP, compared with the previous-generation product.

 Uses Huawei Platinum PSUs with a maximum power conversion efficiency of 94%.

 Supports power capping when it operates in single-system mode.

 Provides Intel

®

Intelligent Power Capability, which allows a processor to be powered on or off based on site requirements.

 Uses low-voltage, low-power E7-8800 v2 processors, suited for data centers and telecommunication environments that have power and thermal limitations.

2.4.3 RH8100 V3 Advantages

Stability and reliability

 Has 60 RAS features, which maximizes server availability.

 Supports memory migration, combined with memory board hot swap, to implement

DIMM replacement without shutdown.

 Complies with component selection criteria for carrier-grade component reliability and derating, and meets carrier-grade reliability test requirements.

 Implements redundancy for key components, such as the basic input/output system

(BIOS) flash memory, iBMC image, PSUs, and fan modules.

Excellent Performance



Supports a maximum of eight E7-8800 v2 or v3 processors. Each processor supports up to 18 cores and 36 threads, and the server supports 144 cores and 288 threads.

 Supports a maximum of 192 DDR3 or DDR4 DIMMs.

Leading Architecture







Uses the Huawei FusionPar technology, and supports hard partitioning and one-click switchover.

Supports flexible configurations based on customer requirements, including LOMs, one or two RAID controller cards, front I/O modules, and four or eight processors.

Adopts modular design for easy upgrades and expansion.

 Stably operates at 40°C (104°F) for a long time.

Simple Maintenance



Adopts modular design for easy maintenance.



39








2 Huawei Solutions

Supports tool-free installation and maintenance.

Supports online parts replacement to facilitate maintenance. The parts include the

DIMM, PCIe card, hard disk, fan module, and PSU.

Uses the latest Huawei iBMC with a Huawei-developed Hi1710 chip to provide overall, one-stop device management, configuration, and fault management.

 Provides a 3.5-inch LCD for onsite maintenance.

2.4.4 RH8100 V3 Technical Specifications

Item

Form factor

Processor

Specifications

8U rack server without holding rails





A maximum of eight Intel

®

Xeon

®

E7-8800 v2 (IvyBridge-

EX) processors. Each processor supports up to 37.5 MB L3 cache capacity and 15 cores. (Four processors can also be configured.)

A maximum of eight Intel

®

Xeon

®

E7-8800 v3 (Haswell-

EX) processors. Each processor supports up to 45 MB L3 cache capacity and 18 cores. (Four processors can also be configured.)

Chipset Intel Patsburg PCH-J (C602-J)

Number of DIMM slots Up to 192 DIMM slots in 16 memory boards(Each processor supports two memory boards.)





E7 v2: supports only DDR3 DIMMs and memory boards.

E7 v3: supports only DDR4 DIMMs and memory boards.

NOTE

192 DIM M slots are supported only when you use the 12-DIM M memory boards.

12 TB (using 64 GB DIMMs) Maximum memory capacity

Number of hard disks Twelve (FM-A), twenty-four (FM-B), or eight (FM-C) 2.5-inch hot-swappable SAS/SATA HDDs or SSDs

RAID support A RAID controller card supports either of the following:



RAID 0, 1, 10, and 1E





RAID 0, 1, 10, 5, 50, 6, and 60, a maximum cache capacity of 2 GB, and a supercapacitor for power-off protection

One or two RAID controller cards of the same model

Network port

PCIe slot

The LOM can be flexibly configured to provide either of the following ports:

 Up to two LOMs



2-port GE-RJ45/4-port GE-RJ45/2-port 10GE SFP+/2-port

10GE RJ45

Supports a maximum of 18 PCIe slots if the server is



40


Item

2 Huawei Solutions

Specifications

configured with FM-A.



Two PCIe 3.0 x16 and eight PCIe 3.0 x8 slots are on the

BM, used for installing standard cards that need to connect to external cables, for example, NICs, HBAs, and IB cards.

 Two PCIe 3.0 x16 and four PCIe 3.0 x8 slots are on FM-A by using two PRMs, used for installing standard cards that do not need to connect to external cables, for example:

 Six Huawei ES3000 PCIe SSD cards

 Two Huawei-certified GPUs

 Three Huawei ES3000s and one GPU

 Two PCIe slots are dedicated for RAID controller cards.

External ports and buttons





Front panel: two USB 2.0 ports, one Power button, one UID button, one VGA port, and one touchable LCD diagnosis panel

Rear panel: Two consoles. Each provides two USB 2.0 ports, one VGA port, one serial port, one GE management network port, one UID indicator, and one LOM slot.

NOTE

Some ports on the two management modules on the rear are unavailable. For details, see 错误!未找到引用源。错误!未找到引用

源。.

DVD-ROM drive

Graphics card

PSU

DVD-RW drive with one SATA port

The graphics card is integrated into an HFC, and supports 32

MB display memory and a maximum resolution of 1280 x

1024.

You can choose 3000 W AC, 2000 W AC, or 2500 W DC

PSUs.

The PSUs can be configured in 2+2 or 1+1 redundancy mode.

System management Supports remote management, WebUIs, virtual KVM, and standard protocols such as IPMI 2.0 and SNMP.

Power-on password and administrator password Security

Supported OSs

Warranty

Windows Server 2012 R2, RHEL 6.5, SLES 11.3, and so on

Huawei provides a three-year warranty for parts replacement and onsite limited repair for the server. Huawei provides 9 x 5

(a 9-hour-a-day, 5-day-a-week) next business day (NBD) support. Optional service upgrades are available.

Dimensions (H x W x D) 352 mm x 447 mm x 855 mm (13.86 in. x 17.60 in. x 33.66 in.)

Weight

Physical environment

Maximum weight: 160 kg (352.80 lb)

Operating environment



Ambient temperature: 5°C to 40°C (41°F to 104°F)

The operating temperature decreases by 1°C (1.8°F) as the



41


Item

2 Huawei Solutions

Specifications

 Ambient humidity: 10% RH to 90% RH (twmax = 29°C)

Storage environment

 altitude increases by 300 m (984.25 ft).

Maximum altitude: 3000 m (9842.52 ft)



Ambient temperature: –40°C to +65°C (–40°F to +149°F)

Ambient humidity: 5% RH to 95% RH (twmax = 38°C)

Acoustic noise The data listed in the following is the declared A-weighted sound power levels (LWAd) and declared average bystander position A-weighted sound pressure levels (LpAm) when the server is operating in a 23°C (73.4°F) ambient environment.

Noise emissions are measured in accordance with ISO 7999

(ECMA 74) and declared in accordance with ISO 9296 (ECMA

109).





Idle:

 LWAd: 6.6 Bels

 LpAm: 50.3 dBA

Operating:

 LWAd: 6.9 Bels

 LpAm: 52.7 dBA

NOTE

The actual sound levels generated during server operating vary depending on the server configuration, loads, and ambient temperature.

2.4.5 RH8100 V3 Server Management

The RH8100 V3 integrates the latest iBMC. iBMC provides the following functions to improve management efficiency and reduce OPEX:







Remote server management and carrier-class reliability

As a Huawei home-grown intelligent management system, iBMC is advanced software for remotely managing servers. It supports KVM redirection, text console redirection, remote virtual media (mapping the CD-ROM drive, floppy disk drive (FDD), and hard disks from the terminal to the server), and IPMI 2.0-based hardware monitoring and management. iBMC is designed based on the carrier-class reliability requirements and supports dual-image backups for software.

Various user interfaces iBMC provides various user interfaces, suc h as the CLI, web-based UI, IPMI management interface. All user interfaces adopt a highly secure encryption algorithm, ensuring access security.

Comprehensive server monitoring iBMC not only monitors servers, but also provides diversified alarms and detailed logs.

For example, logs record processor core temperatures, voltages, fan speed, PSU faults,



42

Huawei SAP HANA Appliance Technical White Paper single node edition 2 Huawei Solutions and bus faults. In addition, iBMC allows you to query the information about processors, memory, and hard disks.



Last-screen information recording and screenshot capturing

When a server about to break down, iBMC saves the last information displayed on the screen for fault identification. iBMC allows you to configure regular or periodical tasks for capturing screenshots, which requires no manual intervention and saves maintenance time. iBMC supports the following features and protocols:























Management interfaces

The server provides IPMI, call-level interface (CLI), Hypertext Transfer Protocol Secure

(HTTPS), Simple Network Management Protocol (SNMP), and Web Service

Management interfaces, meeting various system integration requirements.

Compliance with IPMI 1.5 and IPMI 2.0 iBMC provides standard management interfaces, which allow integration with standard management systems.

Fault detection and alarm management iBMC supports fault detection and alarm management, ensuring stable uninterrupted system operation.

Virtual KVM and virtual media iBMC provides virtual KVM and virtual media, facilitating remote maintenance.

Web user interface (WebUI) iBMC provides the WebUI, helping users to rapidly set and query tasks.

Breakdown screenshot and video iBMC creates screenshots and videos when the system collapses. The screenshots and videos provide clues to the cause of system breakdown.

Screen snapshot and screen video iBMC offers screen snapshots and videos, which simplify routine inspections.

Support for domain name system (DNS) and Lightweight Directory Application Protocol

(LDAP) iBMC supports DNS and LDAP to implement domain management and directory service. This feature simplifies the server management network.

Software dual-image backup iBMC provides software dual-image backup, which allows software to restart from the backup image when a failure occurs. This feature improves system security.

Asset management iBMC implements asset management.

Intelligent power management iBMC allows you to use power capping technology to improve deployment density and uses dynamic power saving technology to reduce OPEX.



43


2.5 ES3000 SSD

Figure 2-18 ES3000 front view

2 Huawei Solutions

The ES3000 V2 is a high-performance PCIe SSD that is designed to eliminate service input/output (I/O) bottlenecks for the Internet, data center, enterprise, and telecom applications.

The ES3000 V2 features high performance, high reliability, low latency, and ease of maintenance, and it applies to services such as databases, search engines, cache acceleration, video acceleration, and high-performance computing (HPC).

The ES3000 V2 is a standard PCIe 2.0 x8 full-height half-length or half-height half-length

SSD, and it can be installed in a universal server.

Performance Features















Device-based architecture: reduces the host resource consumption and the dependence on the host.

Multi-stage multi-channel concurrent processing technology: maximizes the flash granule performance and provides high bandwidth.

Intelligent, adaptive recovery algorithm: effectively reduces performance fluctuation, and makes the device more suitable for service applications such as the database.

Adjustable over-provisioning factor architecture: allows you to customize the visible capacity and expand redundant space to improve write performance of small data blocks, to apply to various application scenarios.

A synchronous command queue technology: concurrently processes a maximu m of 256 read and write commands, which increases concurrence for flash operations and improves input/output operations per second (IOPS).

Direct memory access (DMA) address scatter-gather technology: increases data volume for a read or write request to improve I/O bandwidth.

Native PCIe interface: provides higher bandwidth and effectively reduces I/O latency.

Reliability Features





Enhanced error checking and correcting (ECC) algorithm: provides higher error correction capability than flash granule and ensures device reliability.

Embedded RAID 5 alike algorithm: recovers data after an error occurs in a channel to implement channel-based error correcting.



44




2 Huawei Solutions

Patented flexible RAID algorithm: provides RAID protection for recovered data after a channel fails.



Intelligent wear leveling algorithm: intelligently levels the flash granule wear pressure and prolongs the device service life.





Advanced flash access technology: applies the read retry and adaptive read technologies of flash granules and ensures data validity.

Data inspection technology: periodically inspects data and prevents errors.



Power-off protection: prevents SSD data loss when a power failure occurs on the server.

Simple Management and Maintenance









Provides in-band online upgrades to facilitate routine maintenance.

Provides device information in a centralized manner, including the types, capacities, versions, remaining service life, and health status.

Supports log query for monitoring device health status.

Provides the manufacture dates and serial numbers to facilitate asset management.



45

Huawei SAP HANA Appliance Technical White

Huawei SAP HANA Appliance

Technical White Paper single node Solution

Copyright © Huawei Technologies Co., Ltd. 2016. All rights reserved.

Trademarks and Permissions

Notice

Huawei Technologies Co., Ltd.

About This Document

Contents

SAP HANA

Figure 1-1 SAP HANA architecture

Table 1-1 SAP HANA's innovative technologies

Technology Description

Technology

Description

Figure 1-2 In a traditional relational database management system (RDBMS) or SAP-based

Figure 1-3 SAP HANA data market application

Figure 1-4 New application of SAP HANA

Figure 1-5 SAP HANA BW application

Figure 1-6 Maintenance process of Huawei SAP HANA appliance

DR key indicators

Figure 1-7 PRO and RTO

Backups

Storage Replication

System Replication

Host Auto-Failover

Figure 1-8 SAP HANA HA solution description

Huawei Solutions

Figure 2-1 Huawei RH5885H V3 server-based SAP HANA of high performance configuration

Figure 2-2 Huawei RH5885H V3 server-based SAP HANA of common configuration

Figure 2-3 Huawei RH8100 V3 server-based SAP HANA of high performance configuration

Figure 2-4 Huawei RH8100 V3 server-based SAP HANA of common configuration

Figure 2-5 System Replication logical architecture

Figure 2-6 System Replication solution networking

Figure 2-7 SAP HANA IP address-based redirection

Figure 2-8 SAP HANA DNS-based redirection

Figure 2-9 SAP HANA Auto takeover solution

Figure 2-10 SAP HANA Auto takeover process

Figure 2-11 Failover process of the Host Auto-Failover solution

Figure 2-12 Host Auto-Failover solution networking

Figure 2-13 RH5885H V3 front view

Figure 2-14 RH5885H V3 structure

RAS Features

Performance and Scalability

Manageability and Security

Energy Efficiency

High RAS to Improve Stability and Enable Quick Recovery

Table 2-1 RH5885H V3 RAS features

No. Module Feature Name

Description

No.

Module

Feature Name

Description

No.

Module Feature Name

Description

No.

Module Feature Name

Description

Leading Computing Performance

Table 2-2 RH5885H V3 technical specifications

Item Specifications

Item

Specifications

Item

Specifications

Item

Specifications

Table 2-3 iMana 200 specifications

Item

Specifications

Item

Specifications

Figure 2-15 RH8100 V3 Server view

Figure 2-16 RH8100 V3 components (front view)

Figure 2-17 RH8100 V3 components (rear view)

Huawei FusionPar Technology

Performance and Scalability

Availability and Serviceability