I am pleased to announce the general availability of HA/DR feature for SAP Business Suite on SAP Adaptive Server Enterprise version 16.0 SP02 PL04. This version brings a lot of exciting features that have been highly anticipated.

 

In the new setup, a lot has changed (for the better). Some of the highlights of the new release are as follows:

 

  • Additional replication modes are now supported with SAP ASE 16.0
    • Synchronous Replication Mode (Hot Standby)
    • Near-Synchronous Replication Mode (Hot Standby)
    • Asynchronous Replication Mode (Warm Standby) – Also present in SAP ASE 15.7
  • Only remote distribution mode is supported
  • SAP Replication Server will be installed on the same hardware as SAP ASE 16.0 (co-located). No additional hardware is required.
  • Additional features for monitoring SAP Replication Server using RMA (Replication Management Agent).
  • HA/DR Configuration will be done using a command line utility called “setuphadr” which will make automation a lot easier than in the past.

 

HA/DR Architecture with SAP ASE 16.0 is as follows:

 

 

NW_HADR_Flow.jpg

 

For more information, see:

 

To report this post you need to login first.

5 Comments

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

  1. E Suresh

    Thanks Wajeeh Samdani,


    Can you please more elaborate below.

    • Synchronous Replication Mode (Hot Standby)
    • Near-Synchronous Replication Mode (Hot Standby)
    • Asynchronous Replication Mode (Warm Standby)

    Where can I get installation documentation.

    I checked in installation directory but they have not updated.

    One more question when we use Synchronous/Near – Synchronous modes what is the performance impact on the primary system.

    Regards

    Suresh

    (0) 
    1. Wajeeh Samdani Post author

      Synchronous Replication Mode – ASE 16 only (“Hot Standby” / HA solution)

      • In the synchronous mode, Adaptive Server only commits a transaction after writing the transaction to the primary Adaptive Server database transaction log and after receiving acknowledgement that the replication server has received and written the transaction to SPQ (Simple Persistent Queue).

      Near-Synchronous Replication Mode – ASE 16 only (“Hot Standby” / HA solution)

      • In the near-synchronous mode, Adaptive Server only commits a transaction after writing the transaction to the primary Adaptive Server database transaction log and after receiving acknowledgement that the replication server has received the transaction in memory.

      Asynchronous Replication Mode (“Warm Standby”/ DR solution)

      • Adaptive Server commits a transaction after flushing the transaction to the primary Adaptive Server database transaction log. Whether the replication agent can or will send the transaction to the replication server is independent of the transaction commitment in Adaptive Server.


      Updated documentation “HADR User Guide” will be uploaded to the official documentation page that I linked. Also, you can also find it on SCN at the following link


      I guess when you decide to choose which method to use then you need to evaluate your network latency whether it can support sync replication. If your network latency is very good and temporarily there is a latency increase then replication will switch from sync mode to async mode and switch back to sync as soon as latency returns back to normal. The value for defining what is “high latency” is defined by a parameter which as a default value of 500ms. Which means, if latency goes above 500ms then replication will switch to async and as soon as it comes back down below 500ms, it will switch back to sync mode. This additional functionality was added to avoid any major impact on Primary ASE in the case of temporary latency increases.

      (0) 
  2. Tomasz Romanowski

    I have tested HADR on my virtual environment and I have two extra questions:

    1. Can we use this “stream replications” type (synchronous replications) in customer replication server configuration (eg. MSA or regular replication)? It seems that this should be only RA and RS configuration problem. Can we use synchronous replications in other configuration then HADR?
    2. Based on the graph that you attached, can we separate RS1 and RS2 (from primary and standby ASE) into extra machine? I am asking this because when you want to use standby ASE (in read only mode) for any reports, it is good to limit RS influence into HOST2 (standby env) performance. I am afraid in real world case when RS2 is replicating a lot of real data (it uses HVAR)  it will consume a lot of CPU and RAM. So if we could separate RS into another machine, this could give us almost all standby ASE performance to any report application.
    (0) 
    1. Wajeeh Samdani Post author

      Hi Tomasz,

      1. I am not sure what you mean with MSA or regular replication. In a Business Suite scenario, you can only use this setup for HADR purposes.

      2. Well, in SAP ASE 16.0 with SAP Business Suite setup, it is not supported to separate ASE and SRS.

      (0) 

Leave a Reply