Skip to Content

just came back from our summer meeting of the DSAG data archiving work group and want spread some news.

The HANA hype has hit our company too and while the marketing material reads nice you always stuck when it comes to a detailed question for such a new product.

Today HANA and data archiving was one topic among others, but it was THE topic that got me actually to the DSAG meeting as we had heard some rumors that there will be no data archiving with HANA and wanted some clarification. Further we had no clear picture about Hot and cold storage. Actually we know this term well from our logistic processes in the chemical industry but had some confusion how it is done in SAP.

Hot and cold storage are actually partitions in a database, which is much different to data archiving where you have productive data and archived data usually in 2 different databases

It was said by SAP represenatives:

  • SARA will still be around with HANA Business Suite and S/4HANA, because there will be many business objects around that are not changed and will be archived with SARA transaction as it is today

  • HANA Business Suite will use Data Aging to move older data from the Hot partition to the cold partition in the same database for all changed processes like sFIN (Simple Finance)
  • HANA Business Suite comes with data aging content for SFin, Idoc, Application log, Status management (table JEST)
  • No plans to add further data aging objects to the HANA Business Suite



  • S/4HANA will use Data Aging instead of data archiving for all changed processes
  • Beside of the objects that are mentioned for the HANA Business Suite SAP is working on data aging objects for material documents, purchasing documents, production orders (the core part of Simple Logistics)
  • There is currently no data archiving for the objects which are covered by data aging, for example the BSEG data from Financial document will stay in S/4HANA and cannot be archived until SAP has developed new archiving programs. They are in the early phase of evaluating options and think that it will need some years to a final solution.



  • You can always access the data which was archived before you upgraded to S/4HANA.
  • There are no plans to dispose the SARA transaction




I hope this will clear some doubts that you probably had in common with us

To report this post you need to login first.

11 Comments

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

  1. Matt Fraser

    There was some talk during keynotes at SAPPHIRE about the data aging and hot/cold partition concept. I think it was in Hasso Plattner’s keynote, actually. However, I don’t recall anything about archiving, specifically, and the data aging concept wasn’t presented as any kind of archiving solution but rather as a performance optimization solution.

    (0) 
    1. Jürgen L Post author

      Data Aging is for sure for performance and not seriously meant as a replacement for data archiving. It was just our impression that it is the replacement since there is no data archiving available currently for the new designed products. I am not sure if this impression was so wrong, I guess that the reduction of used table space with S/4HANA in general lead to the assumption that archiving could be a kind of obsolete, especially as you need some historic data to run all those new predictive analysis.

      (0) 
      1. Bala Koteswara Rao Ganduri

        There isn’t any clear road map provided by SAP on this topic as of now.May be we can expect that soon from SAP on the same.

        But for sure I donot think  data archiving will be extinct with HANA.

        (0) 
      2. Matt Fraser

        I would have a hard time imagining that archiving would go away, as it’s not only about about data volume reduction, but also about legal data retention policies. Many organizations really don’t want to have certain data available any longer than the law says they must, as otherwise it becomes ‘discoverable’ during a lawsuit. Business legal departments and records retention groups often advise the destruction of data once it is no longer legally required for retention.

        (0) 
        1. Jürgen L Post author

          This was discussed in the meeting too. While it is more product liability related in the US in Germany we are actually forced by law to destruct data based on the federal data protection act, but this is more related to personal data (HCM system) than the usual ERP data. Based on the current data aging objects there is none where we are forced to do data destruction by law.

          Some members mentioned Groundhog day, said 20 years ago when SAP moved to R/3 it were the same objects that are currently in the data aging, and all others got developed later

          (0) 
  2. Kevin Hill

    Hi,

    Note 2190848 – while dealing with other topics also lists out archiving objects that are obsolete.

    I find it strange that in some areas  the archiving object has been made obsolete and  an data aging object added. The direct implication of no archiving is that this data necessarily needs to remain in the database – and cannot be removed as was possible previously in archiving.

    As more and more archiving objects get replaced with data aging objects, wouldn’t this be a problem from a database size point of view ? – backup times, disk space consumed etc.

    Also I had some thoughts/clarifications/questions around the way data aging would function.

    • If cold data is queried, the relevant partition would be loaded into memory and would remain loaded till 1) there is a memory constraint that would cause this to be unloaded. OR 2) A manual job could unload this ? [ this is my current understanding …not sure if its correct ]
    • Would most report programs by default exclude cold data ?
    • A similar question around custom programs that would query a table that has part of its data as cold – would this cause the relevant partition to be loaded into memory ? – Is there a way a custom program can behave differently when cold data is requested ?
    • When a HANA DB is started, we normally run loadAllTables.py before we release the system. Would this not load the cold partitions ?

    regards,

    Kevin

    (0) 
    1. Jürgen L Post author

      Thanks for the note, I hadn’t seen it yet.

      I am not able to answer any of your questions, we are not yet on HANA with our ERP system and the desired project for next year got already postponed.

      Meanwhile we try to archive as much as possible. Fortunately the biggest part of our system are rolled-in companies who had earlier their own SAP system, they all came just with master data but without history. According to the course details the data in  S4/HANA need much less space (no indices, no redundancies). Altogether we have still some time to look forward to a S4/HANA archiving solution. As indicated in the blog, SAP is working on it, and the pressure is certainly increasing with the number of installations.

      (0) 
    2. Wojciech Lukowski

      Hi,

      Standard unloading mechanism according to LRU approach is still valid for cold partitions. When it comes to keeping cold data under control more actively you might consider a few options:

      – setting up limits for paged memory pool

      – auto-unload feature with retention periods set on partition level

      – standard LRU but perhaps with distributing cold partitions to a dedicated host

      As of SPS 09, load and pre-load do not consider cold partitions.

      I will try to post a separate blog on this to give some more details (hopefully soon).

      Access to cold data from SAP ERP is blocked by default. To access cold data on application level you need to change the session or stack temperature before querying the database. For standard SAP programs (e.g. display account balance, display document, etc) this is already handled. For your custom programs you would need to add a couple lines of code to explicitly request access to cold storage when necessary. You can find more information on this topic in the recently released e-bite https://www.sap-press.com/data-aging-for-sap-business-suite-on-sap-hana_4041/

      BTW: For cold partitions created with paged attribute (recommended for aging) loads do not just happen column-wise but in more granular way i.e. only data that is in the area in which you are searching will be loaded.

      Best regards,

      Wojciech

      (0) 
  3. Kolusu Suresh Kumar

    Hello Jürgen L,


    This is one of the queries that need to respond during our customer workshop.


    While respond to them, still trying to find answer:

    Two of the important advantages that we have with Archiving are:

               1) performance rescue 2) application of legal bounds.


    With hot/warm/cold concept, increased processor cores & GB’s of RAM where is the performance degradation before the data reaches its legal release.


    Does the surviving SARA will still ever be utilized?


    Thanks,

    Kolusu

    (0) 
    1. Jürgen L Post author

      Can’t answer this question. The blog was just a recap of a DSAG session where I was a listener.

      Maybe there is an update in the next Summer meeting of the DSAG archiving group

      (0) 

Leave a Reply