As a Basis Consultants we used to get lot of Requests from End User’s regarding Slowness of BI Systems like Reports Access Taking more time, unable to Access WEBI or Crystal Reports and  rendering issues etc…. this is where Performance Issue comes into Picture.

Sizing is one of the primary tasks that need to be  Taken care by  BI Architects or Technical Consultants While Building Business Intelligence platform.

SAP has Provided many resources for Sizing.

Based on No.of End users, Functional , Technical (Expert ) users we can make Capacity Planning.

  1. For Improved Performance and Scalability we need to Split Adaptive processing Server and Adaptive Job Server. Below note provided Clear process to Split APS and AJS.

https://launchpad.support.sap.com/#/notes/1694041 –> Splitting Adaptive Processing Server

https://launchpad.support.sap.com/#/notes/1950573  —> Splitting Adaptive Job Server

2. Platform Search Application–> Need to change Continuous Crawling to SCHEDULED Crawling

This Application helps users to Search their Documents in BI Launchpad without any performance Issue. Each Document/report that got published /modified/Deleted will be indexed.

In BI 4.2 SP4 Search Algorithm Performance has been Improved

where as SAP  Provided an  Automated tool to perform Capacity Planning i.e. System Configuration Wizard .

System Configuration Wizard is a tool  in SAP BI. we can Access that through CMC. This wizard Guides us through initial Configuration steps required to our BI Landscape.

  1. Capacity
  2. Services Auto Start
  3. Folders Path

once After Completion of  Installation and After Logging into CMC, System Configuration Wizard Screen Pop’s up.

Capacity Planning Plays vital Role BI Environment. Based on RAM Size  in Our System, Number Of Adaptive Processing Server’s are Deployed.

System Configuration Wizard Screen:

Based on RAM Size, Below are Recommendations provided by SAP

6-8GB —–>1 APS Recommended (XS)

12-16GB—>4 APS Recommended (S)

15-25GB—>7 APS per Node (M)

30-45GB—>9 APS per Node (L)

40-60GB—->11 APS per Node (XL)

we can Access System Configuration Wizard from CMC.

We can Specify Folder paths like Input File Repository, Output File Repository, Log, Data, Auditing etc..

 

References:

https://launchpad.support.sap.com/#/notes/1735129/E ——-> Sizing Estimator

https://launchpad.support.sap.com/#/notes/2322354/E —>Deployment and Sizing Best Practices

https://archive.sap.com/documents/docs/DOC-33183  —>  Resources related to Sizing are provided in this blog

https://blogs.sap.com/2017/04/11/sap-bi-4.2-sp4-preview-of-whats-new-in-web-intelligence-and-semantic-layer/

 

To report this post you need to login first.

1 Comment

You must be Logged on to comment or reply to a post.

  1. Moritz Hödel

    Hi,

    I disagree that the AJS has to be split (and it’s even not recommended to do so).
    Even the Note you are referring to is stating exactly this in the cause section:

    • The Adaptive Job server(AJS) is not a processing server and is more like a launching pad (starting other processes), hence its not necessary to Split the Adaptive Job Server

    Actually this can lead to several issues.

    And if you mention how to fix the performance issues caused by the Platform Search settings,
    you should at least tell the whole story 🙂
    Scheduled crawling is just the first step, even if it’s the most important imo.
    You have missed to tell about unchecking Universes, switching the level of indexing, skipping instances, clustered deployments, etc. …
    I would recommend to have a look at these:

    Note 2241459 – *** MASTER KBA *** Platform Search Application

    Note 1863726 – How does BI 4.0/4.1 Platform Search Application indexing work?

    Behind the Scenes with Platform Search in Business Intelligence Platform 4.x

    Note 1717717 – How to troubleshoot high CPU consumption issues on BI 4.0.

    Note 2183804 – Configurable option to skip instances from Platform Search indexing

    Note 1844013 – How to configure Platform Search Index location in clustered environment?

    (1) 

Leave a Reply