SAP 2

SAP HANA Sizing

SAP HANA Pre OS Installation Sizing:

Sizing for this categories is needed:

1. memory for static data
2. memory for objects created during runtime
3. disk sizing
4. CPU sizing

RAM

SDF (Source Data Footprint) is current Oracle DB size without indexes and blobs (for Oracle use dba_segments).

RAM size = SDF * 2 / 7 (i.e. RAM Size = 300 * 2 / 7 = 85GB but HANA minimum recomendation is 128GB of RAM)

Describing formula:

1. Both static (in-memory store) and dynamic (executing queries and loading data) RAM data sizes are sized. "2"  comes from RAM dynamic = RAM static.
2. "7" comes from average compression factor database table size : HANA memory = 7 : 1

HDD

Backup storage >= /hana/data + /hana/log

• OS = 50GB 
• /hana/data  = 1 * SDF (i.e. Disk data = 1 * 300 = 300GB)
• /hana/shared min 1*RAM and up to 1TB = 85GB
• /hana/log (1/2 * 85 = 42.5GB):
• if RAM Size <= 512GB = 1/2 * RAM
• if RAM Size > 512GB = 512GB
• /usr/sap = 50gb
• Total HDD = 50 + 300 + 85 + 42.5 + 50 = 527.5GB

CPU

Approximation:

0.0625 core/vCPU is needed per 1GB of RAM

core/vCPU =0.0625 * 128 = 8

Sizing summary

SDF = 300GB

RAM = 128GB

HDD = 527.5GB

core/vCPU = 8

SAP 1

What is SAP

SAP ERP (Enterprise Resource Planning) software is central application which covers Sales and Distribution (SAP SD), Production Planning (SAP PP), Financials (SAP FI), Human Capital Management (SAP HCM) and many more.

SAP ERP is a transactional system with data being written to its database frequently in a productive business environment.

Analytical operations are usually performed in SAP BW or HANA models.

SAP BW (Business Warehouse) - data warehouses (DW) are applications on a separate database that take data from source systems, then using ETL (Extraction, Transformation, Loading) process cleanse it, apply business logic and optimally store it within themselves for faster access from reports. ETL process is used to more quickly access KPIs (Key Performance Indicators - are being measured across time frames and used for measuring companies to make forecast and decisions) of the company.

Before SAP BW on HANA arrived, one of the biggest downfalls of SAP BW was that there were too many layers of redundant data from pre-cleansing, cleansing and multiple layers of transforming data. All this inherent latency (time taken for data to reach from source to reporting layer) made it a bulky truck which always got the job done but was never meant for a street race. Then came SAP HANA – SAP’s new revolutionary database and then the idea to take the BW application off the traditional databases and put it on top of HANA.

And now BW on HANA is BW4/HANA

SAP flow is as follows:
1) ERP
2) Cleansing and homogenization of data
3) BI
4) from BI to Business Strategy
5) ERP

Calculating ASA Throughput

1. copy output of command below to the asaThroughput text file:
1. show traffic | begin Aggregated Traffic on Physical Interface
2. delete all traffic statistics for interfaces with "Internal" word
3. Automatically count bytes in all interfaces statistics:
1. bytes=$(grep -E "bytes$" asaThroughput  | \
2. grep -v "0 packets.*0 bytes" | \
3. awk '{s=s+$3} END {print s}')
4. select the first appearance of the number of seconds of traffic statistics:
1. seconds=$(grep -m 1 secs asaThroughput | \
2. awk '{print $3}' | \
3. cut -d\. -f 1)
5. get final result (ASA throughput in Mbps):
1. echo $((bytes/seconds*8/1024/1024))

DSL chip codes/abbreviations

ALCB Alcatel
ANDV Analog Devices
BDCM Broadcom
GSPN Globespan
IKNS Ikanos
IFTN Infineon
META Metanoia
STMI STMicroelectronics
TSTS Texas Instruments