Home > Workload Solutions > SAP > Guides > Dell Solution for SAP Landscape Consolidation > Sizing example
The following sizing example uses sample customer inputs to show how you can determine compute and storage requirements for SAP landscapes.
A customer has three PRD systems, ERP, CRM, and BW, each with a system landscape consisting of DEV, QAS, and SBX systems. Based on its Quick Sizer projects, the customer provided the information in the following table or each PRD system.
Application | Application type | Total SAPS | Disk size (GB) |
ERP | OLTP | 192,000 | 8,000 |
CRM | OLTP | 64,000 | 2,000 |
BW | OLAP | 120,000 | 4,000 |
The customer has two data centers that are 30 miles apart and has chosen to use VMware virtualization for server consolidation and high availability.
The customer advised that the quality system should be sized the same as production, with a VMware CPU overcommit factor of 2. Both the DEV and SBX systems should be sized at 50 percent of production, with a VMware CPU overcommit factor of 4.
SRDF/Metro is included for storage continuous availability and disaster recovery with storage replication for the PRD and QAS systems. A Data Domain system is included for data protection and backups. The annual growth forecast is 7 percent with a maintenance term of 4 years.
The following table shows the SAP landscapes that the Dell SAP team derived from the customer inputs:
System | Application | Type | Description | Site | Replication site | SAPS (input) | Disk size (input)GB | Annual disk growth (input) |
BM1 | ERP | OLTP | PRD | A | B | 192,000 | 8,000 | 7 % |
QM1 | ERP | OLTP | QAS | B | A | 192,000 | 8,000 | 7 % |
DM1 | ERP | OLTP | DEV | A | — | 96,000 | 4,000 | 7 % |
SBX | ERP | OLTP | SBX | B | — | 96,000 | 4,000 | 7 % |
CRM | CRM | OLTP | PRD | B | A | 64,000 | 2,000 | 7 % |
CRQ | CRM | OLTP | QAS | A | B | 64,000 | 2,000 | 7 % |
CRD | CRM | OLTP | DEV | B | — | 32,000 | 1000 | 7 % |
CRS | CRM | OLTP | SBX | A | — | 32,000 | 1000 | 7 % |
BWP | BW | OLAP | PRD | A | B | 120,000 | 4,000 | 7 % |
BWQ | BW | OLAP | QAS | B | A | 120,000 | 4,000 | 7 % |
BWD | BW | OLAP | DEV | A | — | 60,000 | 2,000 | 7 % |
BWS | BW | OLAP | SBX | B | — | 60,000 | 2,000 | 7 % |
Next, we used the customer inputs from Table 12 to calculate the compute and memory requirements, as shown in the following table:
System | Site | SAPS (input) | Database | Application SAPS1 | VMware CPU overcommit factor | Database SAPS (with VMware overcommit) 2 | Application SAPS | Database memory (GB)3 | Application memory (GB)3 | Number of application servers1 |
BM1 | A | 192,000 | 38,400 | 153,600 | 1 | 38,400 | 153,600 | 115 | 461 | 5 |
DM1 | A | 96,000 | 19,200 | 76,800 | 4 | 4,800 | 19,200 | 58 | 230 | 2 |
CRQ | A | 64,000 | 12,800 | 51,200 | 2 | 6,400 | 25,600 | 38 | 154 | 3 |
CRS | A | 32,000 | 6,400 | 25,600 | 4 | 1,600 | 6,400 | 19 | 77 | 1 |
BWP | A | 120,000 | 40,000 | 80,000 | 1 | 40,000 | 80,000 | 120 | 240 | 3 |
BWD | A | 60,000 | 20,000 | 40,000 | 4 | 5,000 | 10,000 | 60 | 120 | 1 |
QM1 | B | 192,000 | 38,400 | 153,600 | 2 | 19,200 | 76,800 | 115 | 461 | 5 |
SBX | B | 96,000 | 19,200 | 76,800 | 4 | 4,800 | 19,200 | 58 | 230 | 2 |
CRM | B | 64,000 | 12,800 | 51,200 | 1 | 12,800 | 51,200 | 38 | 154 | 3 |
CRD | B | 32,000 | 6,400 | 25,600 | 4 | 1,600 | 6,400 | 19 | 77 | 1 |
BWQ | B | 120,000 | 40,000 | 80,000 | 2 | 20,000 | 40,000 | 120 | 240 | 3 |
BWS | B | 60,000 | 20,000 | 40,000 | 4 | 5,000 | 10,000 | 60 | 120 | 1 |
1 Ideally, the customer provides a breakdown of the database and application SAPS from a Quick Sizer project. Otherwise, the Dell SAP engineering team applies a general rule to calculate the ratio of database SAPS to application SAPS using total SAPS. For OLTP, the team uses 1:4, and 1.2 for OLAP.
2 The database and application SAPS values for each system are adjusted to include the VMware CPU overcommit factor.
3 To calculate the required memory for each database and application server, we use the general rule for throughput-based sizing, where we estimate 3 GB of memory for each 1,000 SAPS. The team calculates this memory against the SAPS values without the CPU overcommit factor.
Note: In virtual environments, do not overcommit memory for SAP PRD and non-PRD systems.
For each site that is described in the preceding table, the team calculated the total SAPS and memory requirements, as shown in the following table.
Site | Total database SAPS | Total application SAPS | Total SAPS | Total database memory (GB) | Total application memory (GB) | Total memory (GB) |
A | 96,200 | 294,800 | 391,000 | 410 | 1,282 | 1,692 |
B | 63,400 | 203,600 | 267,000 | 410 | 1,282 | 1,692 |
To calculate the number of physical servers that are required, the team compared the total compute requirements for each site with the physical capabilities of the SAPS value and the memory of the chosen server models.
Additional compute resources for high availability and disaster recovery are included on each site. That way, if a server or site is lost, the remaining site has the resources to take over the failed sites workloads.
In this example, the result was calculated using both the R960 and R760 models. The following table shows the physical compute requirements for the solution using the R840 server on each site.
Site | Total SAPS required | Total memory required (GB) | Virtual SAPS | Cores (up to 4 x 60) | RAM (GB)1 | Number of servers SAPS bound2 | Number of servers memory bound3 | Number of servers required4 | Number of servers required with HA5 |
A | 391,000 | 1,692 | 748,017 | 240 | 2,048 | 1 | 1 | 1 | 2 |
B | 267,000 | 1,692 | 748,017 | 240 | 2,048 | 1 | 1 | 1 | 2 |
1 This sizing example uses PowerEdge R840 servers with 768 GB of RAM.
2 The PowerEdge R960 has the capabilities of up to 748,017 virtual SAPS, and the SAPS requirement in the example is 391,000 for Site A and 267.000 for Site B. Therefore, one server on each site is needed to satisfy the requirement.
3 The PowerEdge R860 server in the example has 2,048 GB of RAM and the compute requirement is 2,048 GB of RAM on both site A and site B. Therefore, two servers on each site are needed to satisfy the requirement.
4 The higher value of 2 and 3 is the number of servers required.
5 The Dell SAP engineering team added a server to the value of 4 for high availability.
The following table shows the physical compute requirements for the solution using the R760 server on each site:
Site | Total SAPS required | Total memory required | Virtual SAPS | Cores | RAM | Number of servers SAPS bound2 | Number of servers memory bound3 | Number of servers required4 | Number of servers required with HA5 |
A | 391,000 | 1,692 | 385,020 | 112 | 1,024 | 2 | 2 | 3 | 3 |
B | 267,000 | 1,692 | 385,020 | 112 | 1,024 | 1 | 2 | 2 | 3 |
1 This sizing example uses PowerEdge R760 servers with 1,024 GB of RAM.
2 The R760 server has the capabilities of up to 385,020 virtual SAPS. The SAPS requirement in the example is 391,000 for Site A and 267,000 for Site B; therefore, two servers are required on site A and one server on site B.
3 The R740 server in the example has 1,024 GB of RAM. The compute requirement is 1,692 GB of RAM on both site A and site B, so three servers are required.
4 The higher value of 2 and 3 is the number of servers required.
5 The Dell SAP team added a server to the value of 4 for HA.
The following table shows the storage requirements for Site A and Site B:
System | Site | Replication site | Disk size input (GB) | Database disk final size | Database SAPS (with VMware overcommit) | IOPS (calculated)2 | Number of app servers | Total OS size (GB)3 | Total exe size (GB)3 | Dump size | Total storage requirements5 |
BM1 | A | B | 8,000 | 10,486 | 38,400 | 23,040 | 5 | 768 | 768 | 10,486 | 22,509 |
QM1 | B | A | 8,000 | 10,486 | 19,200 | 11,520 | 5 | 768 | 768 | 10,486 | 22,509 |
DM1 | A | - | 4,000 | 5,243 | 4,800 | 2,880 | 2 | 384 | 384 | 5,243 | 11,254 |
SBX | B | - | 4,000 | 5,243 | 4,800 | 2,880 | 2 | 384 | 384 | 5,243 | 11,254 |
CRM | B | A | 2,000 | 2,622 | 12,800 | 7,680 | 3 | 512 | 512 | 2,622 | 6,267 |
CRQ | A | B | 2,000 | 2,622 | 6,400 | 3,840 | 3 | 512 | 512 | 2,622 | 6,267 |
CRD | B | - | 1,000 | 1,311 | 1,600 | 960 | 1 | 256 | 256 | 1,311 | 3,134 |
CRS | A | - | 1,000 | 1,311 | 1,600 | 960 | 1 | 256 | 256 | 1,311 | 3,134 |
BWP | A | B | 4,000 | 5,243 | 40,000 | 36,000 | 3 | 512 | 512 | 5,243 | 11,510 |
BWQ | B | A | 4,000 | 5,243 | 20,000 | 18,000 | 3 | 512 | 512 | 5,243 | 11,510 |
BWD | A | - | 2,000 | 2,622 | 5,000 | 4,500 | 1 | 256 | 256 | 2,622 | 5,755 |
BWS | B | - | 2,000 | 2,622 | 5,000 | 4,500 | 1 | 256 | 256 | 2,622 | 5,755 |
1 Future database disk size = disk size x annual growth of 7 percent over 4 years
2 To estimate the required IOPS, calculate the database SAPs based on the analytical or transactional consumption of SAPS. See Sizing the disk and disk I/O on page 44. Disk I/O for OLTP applications = 60 percent of database SAPS. Disk I/O for OLAP applications = 90 percent of database SAPS.
3 A disk capacity of 128 GB for the operating system and 128 GB for exe on each system database and application server
4 Storage capacity for a database dump, exports, or logs
5 Total system storage is the value of the disk + operating system + exe columns added.
With storage replication and SRDF/Metro across both sites, the systems being replicated must also be included to size the storage on each site.
The following table shows the total storage requirements for Site A. The total represents requirements for each SAP system residing primarily on Site A as well as the systems from Site B that are being replicated to Site A.
Site A | Storage (GB) | IOPS |
SAP systems | 60,429 | 75,720 |
Replicated SAP systems | 40,286 | 37,200 |
Total Site A requirements | 100,716 | 112,920 |
Using the total storage and total IOPS, the Dell representatives use the PowerMax Sizer tool to provide the configuration of the PowerMax array for Site A.
The following table shows the total storage requirements for Site B:
Site B | Storage (GB) | IOPS |
SAP systems | 60,429 | 45,540 |
Replicated SAP systems | 40,286 | 62,880 |
Total Site B requirements | 100,716 | 108,420 |
Using the total storage and total IOPs, your Dell representatives use the PowerMax Sizer tool to provide a detailed configuration of the PowerMax array for Site B.