Essay PreviewMore ↓
The 5 key components of the new platform for ESAP applications are described below. All together, these components enable the delivery of much more flexible, resilient, highly scalable and performant infrastructure that can support ESAP and other applications with substantially decreased TCO.
22.214.171.124. 64-bit Linux
Linux is commonly recognised as a most cost-effective Open Source operating system for business-critical applications. This platform allows to directly address the large memory, facilitating the consolidation of non-production (development and test) application instances with very low CPU consumption but relatively high memory utilisation without using special and rather expensive server virtualisation software. Linux platform is sufficiently mature, it is less vulnerable to the security threats than Windows, and allows to use SAP Adaptive Computing (see item 5) that is not fully supported on Windows platform. Finally, Linux has native support for NFS protocol, which is crucial to enable the efficient use of NAS. This is a key factor driving the choice of Linux platform instead of competing MS Windows. It allows using NAS directly connected to the Linux servers without iSCSI overhead that would be non-avoidable in the case if Windows platform was used.
126.96.36.199. AMD Opteron
Despite Linux can perfectly run on 64-bit Intel servers as well, AMD provides better performance/cost ratio and is more mature platform for 64-bit computing. Linux on AMD is fully certified by SAP platform for SAP applications. ESAP runs on 64-bit Red Hat Enterprise Linux 4.0 Update 5.
188.8.131.52. Oracle Real Application Clusters (RAC)
Oracle RAC is the database clustering solution which can run on commodity platform without any special requirements to the hardware or operating system software. All the nodes of cluster are operating concurrently with automatic load balancing. The load is automatically redistributed without database services outage in the case of the cluster node failure or the new node introduction. This ensures high resilience and virtually unlimited horizontal scalability of the database layer using small cheap unified commodity servers. Oracle RAC is fully certified by SAP as a DB platform for SAP applications.
184.108.40.206. NetApp storage
NetApp provides fully certified by SAP and Oracle NAS solution which can not be treated as a commodity, but is very widely used for business-critical applications and has some unique features not available from any other storage vendor. Particularly, FlexClone allows sharing the copy of production DB between several non-production instances. This copy may be represented by maintained in the real-time at the failover site remote replica of the production DB, making the DR almost instant (realistically, within 3 hours if dedicated DR hardware is not available and failover happens to the non-production processing environment).
How to Cite this Page
"Key Components of ESAP Technical Platform." 123HelpMe.com. 09 Dec 2019
Need Writing Help?
Get feedback on grammar, clarity, concision and logic instantly.Check your paper »
- There are several components for a successful leader described in the book Be-Know-Do. The book title itself, Be-Know-Do, gives us the three of the main components; each word standing for a different component needed to be a leader. Be is for a leader’s character, Know is the skills the leader possess, and Do is the leader bringing those two together and acting on them. The first principle is Be: The first component in a successful leader is their characters, their mental and moral qualities which are present at all times.... [tags: Leadership, Skill, Problem solving, Management]
847 words (2.4 pages)
- According to Dryden et al. (1992), the concept of art therapy is essentially based in the image and the progression of this method permits the occurrence of a transaction between three critical components of this model. Respectively these components are known as the creator or patient, the artifact, and the therapist. The aims of art therapy align with other processes of therapy whereby; the therapist engages his/her expertise to allow the patient’s unconscious feelings to travel to a conscious state.... [tags: therapy, components, emotions, feelings]
2735 words (7.8 pages)
- ... Reading is not an easy task, but oral language does help with the process. Additionally, oral language would also help with the recognizing and association of words to text that is being read. There may be a situation in which the student is reading about for which they can relate too. This could be due to their prior oral language development. Associating words that are recognized in their vocabulary with words that are in the text creates a link that the student can expand on. This would give the student motivation to read because they are relating and understanding what they are reading.... [tags: phonological, comprehension, fluency]
1680 words (4.8 pages)
- The key components of a charity can be categorised as fundraising money, whether by collecting donations, selling charitable products and distributing aid in all forms and locations. Sense and respond involves determining problems and developing systems to solve such problems (Stair, et al., 2015), which can be extremely useful. The types of system which are beneficial for organisations are operational information systems, management information systems, decision support systems and knowledge support systems.... [tags: Decision support system, Decision theory]
1528 words (4.4 pages)
- Setting and Achieving Goals Short and long term planning are key components of a successful business. Organizations should look at planning as a multitier process. In the long term they will use strategic planning that will focus on the future of the company. In the short term they will utilize operational planning which will implement the strategic plan of the business. Also, through action plans the goals are measurable allowing staff the opportunity to compare progress to expected results. Both short and long term planning are required to meet the goals of the company as they move into the future.... [tags: Management, Strategy, Organization]
716 words (2 pages)
- It is vital that nurses do not let interactions with a culturally diverse patient be influenced by stereotypes. This will place a wall between the nurse and the patient preventing both from understanding one another and developing a trust that is needed in the relationship (Kersey-Matusiak, 2012). The purpose of this post is to review cultural diverse assessments, describe three key components of a comprehensive cultural assessment, and to identify five communication skills that facilitate culturally competent health care.... [tags: Sociology, Health care, Nonverbal communication]
753 words (2.2 pages)
- The ability to design and critically evaluate unit plans and resources is a vital part of the history teachers’ role (Kriewaldt & Fahey, 2012). To do this successfully, teachers must have knowledge of the key curriculum components that are essential for the effective teaching of history (Fahey, 2012). These components include the Australian Curriculum: History (hereafter AC:H) requirements, historical inquiry, quality resources and community partnerships. This essay will discuss the findings of an inquiry undertaken to critically analyse and evaluate a history unit plan designed for Year 4 students in relation to how effectively these components have been incorporated.... [tags: Education, Teacher, Knowledge, The Unit]
906 words (2.6 pages)
- Identifying features and key components of unknown chemical mixtures are an essential part of chemistry. In many cases it is necessary to analyses and understand the chemical makeup of pharmaceuticals. Qualitative analysis is a method of testing a chemical mixture to determine its chemical makeup, such as its functional group. Identifying functional groups of an unknown compound are important, as it helps in understanding how a chemical acts under certain conditions, such as when changing physical states.... [tags: Alcohol, Functional group, Ketone, Carboxylic acid]
2381 words (6.8 pages)
- ... Which is the shipper and the carrier. The shipper is the one that moves the product between two points in the supply chain, while the carrier is that one that moves or transports the product Challenges Transportation System Faces In this day and age, there are many challenges that transport systems faces, it is due to many factors, such as global competition, government budget constraints, increased demand from groups that has special interests, like senior citizens, infrastructure challenges, sustainability issues and energy costs.... [tags: critical roles, key components]
1980 words (5.7 pages)
- Technical communication is most likely going to be the most difficult part of my professional career as a computer scientist. Writing for this field is similar to other fields since programmers are expected to be able to write documents that are easily understandable by the software users, and are readily usable with technical details for professionals. Programmers must write many documents directed at the non-technical audience which consists of text within programs, read me files, help files, manuals on software use, presentations for clients, and many more.... [tags: Technical Writing]
679 words (1.9 pages)
The common concern regarding the non-suitability of NAS storage for large transactional database servers due to the limited bandwidth available with 1 Gbit Ethernet connections is addressed by using small cheap servers for RAC nodes. These individual servers are not big enough to generate the traffic able to saturate 1 Gbit network link, and every server has its own dedicated resilient network connections with NAS storage. The total aggregated connection bandwidth between Oracle RAC servers and NAS may be very high.
220.127.116.11. Adaptive Computing Controller
SAP Adaptive Computing Controller is a special component of SAP technology stack NetWeaver, allowing to dynamically change the personality of the servers and to re-allocate the processing resources on fly, without application outage. It is achieved by running the applications from network-based filesystems, and allows to completely change the purpose of the servers by mounting different filesystem and starting another application, stored on this filesystem. The transparency of this switchover for end-users is achieved via grace shutdown, stopping the initiation of the new user connections with the server and switching over after grace period (usually several hours). This re-allocation of resources can be automated using SAP Solution Manager, or can be managed manually. SAP Adaptive Computing is supported on Unix and Linux platforms.
Unfortunately, even the latest version of SAP Adaptive Computing Controller (SAP ACC) 7.1 doesn’t support some key distinctive features of ESAP platform, like Oracle RAC, MCOD and HA configuration. Due to this the special tools based on the same concepts as SAP ACC, but free of these limitations, were developed by XE IM, enabling the efficient administration of ESAP application instances.
2.3.3. Key distinctive features of ESAP architecture
The main business benefits of ESAP platform couldn’t be achieved without several key distinctive features of ESAP architecture.
One of them is virtualisation. All SAP application services are running as virtual hosts and can be easily moved from one physical server to another, that may be required for example in the case of unrecoverable hardware failure, to optimise the utilisation of hardware resources or for disaster recovery. Every ESAP service like Dialog Instance or Oracle RAC node has a pre-assigned Virtual IP (VIP) address and port number. All application resources are stored on NAS storage and can be mounted on any server, no application-specific data is stored on the local disks of the servers.
Another feature is consolidation, which has two main aspects. First one is the use of MCOD — Multiple Components in One Database. With MCOD, multiple SAP applications share the same instance of Oracle DB, improving the utilisation of storage and minimising the administration costs. Every application is installed in its own DB schema, insuring complete independency of them while served by the same DB instance. The second is running multiple SAP application services and DB instances on the same box, improving the utilisation of processing resources. For small number of concurrent users the whole layer of ESAP landscape may run just on one physical server, delivering full functionality of this very complex application.
Third feature, perhaps one of the most important, is scalability. The design of ESAP implies no limits on the number of SAP Application Servers or Oracle RAC nodes, allowing to scale out the application from just one to hundreds and thousands servers running similar numbers of application instances and RAC nodes, supporting hundreds of thousands concurrent users if required. The capacity of NAS filers (storage controllers) is limited to 96-300 TB, but additional filers can be installed if required and storage connected as additional mount points. The architecture of ESAP allows scaling the server infrastructure up and down without outage of applications.
2.3.4. Resilience of ESAP platform
The traditional means of achieving high availability usually result in a significant increase of the infrastructure costs due to the procurement of expensive high availability hardware and software, or investment into the redundant hardware which runs in idle, delivering no added value during normal operations. It was not acceptable for ESAP, and special attention was paid to ensure that redundant components are active all the time, delivering during normal operations spare capacity that helps to handle peaks of the load. For example, in Production environment every ESAP application has at least 2 instances of each service to enable continuous operations if one of instances will fail. More than 2 instances are required when the capacity of 2 instances is not sufficient to handle the required levels of load. As many as necessary dialog instances are deployed to handle the actual number of concurrent users for any ESAP application. The automated load balancing and re-direction of the new calls in the case of the failure to the surviving instances is a native feature of SAP applications, and it is fully utilised in ESAP design.
With Oracle RAC (Real Application Clusters) a similar concept is supported by the data base server nodes. The database load is automatically distributed across all the RAC nodes – members of the cluster, acting from SAP application perspective as one virtual database server instance, and is re-distributed in the case of any single DB node failure. Only the transactions that perform DB operations using the failed node at the moment of the failure will be cancelled and rolled back.
The resilient design was applied to the network infrastructure, with dual path network using 2 sets of network switches and routers. Almost every single logical network connection is implemented with two physical network interfaces, and these redundant links are bonded, doubling the available bandwidth.
NAS storage is deployed in High-Availability configuration with clustered filers (processing devices that manage NAS I/O operations). Both filer heads are normally active, delivering more spare capacity. It is important that not more than 50% of each filer’s processing capacity is in use to ensure that in the case of one filer failure, another has sufficient spare capacity to take the full load. NAS storage is using RAID 4 (dedicated parity drive) with dual parity. Dual parity ensures that fault-tolerance is preserved in the case when 1 HDD in array is failing, so the array can survive another drive failure without data loss. However, normally the failed drive will be replaced quickly, well before the failure of the second drive.
The single points of failure were eliminated not only in the physical design, but in the logical design of Production ESAP applications as well. At least 2 instances of each type of service are available for every application with automatic load balancing between them and failover in the case of failure. The natural single points of the failure (components that balance the load and despatch the calls) like Message Server, Enqueue Server, Update Service, Web Dispatcher, are implemented in High Availability mode, controlled by Oracle Clusterware scripts. The components that require to store some data in the memory (Enqueue Server) are implemented with constant real-time replication of the data to the failover instance, installed on another physical box and being inactive until failover. One instance of each service is installed on each of 2 boxes. Each ESAP service has individual virtual IP address (VIP) assigned, that allows to move the services from one physical box to another by unbinding the IP address from the old instance of the service and binding it to the new one.
If running primary instance of the service fails, VIP address is moved by Clusterware to the secondary instance of the service on another box and this service is started, taking the processing load.
2.4. Backup and recovery strategy for ESAP platform
Backup and recovery strategy of ESAP applications is based on the advanced features of NetApp NAS. All the critical data volumes are replicated from the site of primary location of this volume to the secondary site.
The replication of data between sites is handled by the NetApp NAS storage feature named FlexMirror, which transfers the updated blocks of data from one site to another. These transfers are done by the storage itself and require no additional hardware resources.
Highly volatile data (redologs) is replicated from primary site to the secondary every 5 minutes for production volumes, and every 6 hours for non-production volumes.
Volatile data (data tables) is replicated from primary site to the secondary every 30 minutes for production volumes, and every hour for non-production (see more details in section 2.10.2).
Such schedule allows in the case of disaster to recover the data at the secondary site with 5 minutes data loss. The data tables before recovery may loose up to 30 mins of data, so the changes recorded in the redologs are applied to the data tables up to the last consistent log file which may still miss the last 5 minutes of data updated after last replication but before disaster.
Consistent snapshots, representing consistent recovery points, are created regularly and replicated to the secondary location. Before creation of the snapshot, database is switched to hot backup mode, so the changes are applied to the log files but not to the data tables. Snapshot is made on the data tables, so the data is consistent. The creation of snapshot takes only several seconds, as it represents only the list of blocks used on the volume. After creation of snapshot database is switched back to normal mode and all the changes saved in the log files are applied to the data tables.
The recovery from snapshot is almost instant. Snapshots can be used as a read-only dataset for backup purposes, or converted into writable FlexClone for DB and application instances cloning. DB consistency check can be run on read-only snapshots for early discovery of any logical data corruption, but it is associated with significant load on storage I/O channels and CPUs as actually every block of DB is being read and validated.
Snapshots are made every 6 hours and retained for 48 hours; daily snapshots are retained for 30 days in production volumes and for 7 days in non-production.
Daily tape backups, stored off-site, are produced for production volumes, with full backup during weekend and incremental backups every night. The tape backups for non-production volumes are created on request.