WebSphere Application Server & Liberty

 View Only
  • 1.  Disaster recovery for websphere

    Posted 20 days ago

    How to sync websphere dr site with main site automatically?



    ------------------------------
    karim mahgoub
    ------------------------------


  • 2.  RE: Disaster recovery for websphere

    IBM Champion
    Posted 20 days ago
    Hi,

    There no such thing natively, it's more an architecture decision. You can opt to have a single DMGR managing instances on both sites (not recommended). So it depends on the use case, but there is command to sync configuration between to separate cells (prod and dr).

    Hope it helps

    Best regards, 
    --------------------------------------------------------------
    Walid Largou
    Solution Architect / Service Delivery Manager
    Mobile : +212 621 31 98 71
    Email : l.walid@powerm.ma
    320 Bd Zertouni 6th Floor, Casablanca, Morocco
    --------------------------------------------------------------
    https://www.powerm.ma
    Power Maroc #1 IBM Business Partner – Power Maroc
    This message is confidential .Its contents do not constitute a commitment by Power Maroc S.A.R.L except where provided for in a written agreement between you and Power Maroc S.A.R.L. Any authorized disclosure, use or dissemination, either whole or partial, is prohibited. If you are not the intended recipient of the message, please notify the sender immediately.





  • 3.  RE: Disaster recovery for websphere

    Posted 20 days ago
    Hi walid,

    Thanks a lot for your support.

    what command can sync main site with DR site ?

    Thanks a lot





  • 4.  RE: Disaster recovery for websphere

    IBM Champion
    Posted 20 days ago
    Hi,

    I think an example is worth a thousand words ��.

    Check this video : https://youtu.be/EXAfSx1gVwc


    Hope it helps

    Best regards, 
    --------------------------------------------------------------
    Walid Largou
    Solution Architect / Service Delivery Manager
    Mobile : +212 621 31 98 71
    Email : l.walid@powerm.ma
    320 Bd Zertouni 6th Floor, Casablanca, Morocco
    --------------------------------------------------------------
    https://www.powerm.ma
    Power Maroc #1 IBM Business Partner – Power Maroc
    This message is confidential .Its contents do not constitute a commitment by Power Maroc S.A.R.L except where provided for in a written agreement between you and Power Maroc S.A.R.L. Any authorized disclosure, use or dissemination, either whole or partial, is prohibited. If you are not the intended recipient of the message, please notify the sender immediately.





  • 5.  RE: Disaster recovery for websphere

    Posted 20 days ago
    Thanks a lot for your support.

    are there any other way to perform automatic sync between dr and main sites ?

    Thanks a lot





  • 6.  RE: Disaster recovery for websphere

    Posted 11 days ago

    @karim mahgoub Carefully consider the impact of automatic synchronization if a change is made to the primary site, a change with a mistake, a mistake which breaks something, and that change is replicated to the backup site, breaking it as well, now both the primary and backup site are BROKEN, and the purpose for having a backup was to avoid this very scenario (and no, this isn't a theoretical example, I've seen it occur)   

    For this and other reasons I other means for maintaining consistency between sites, and while the method cited by @Largou walid will work, it can result in the scenario I cited above is one isn't careful I'll add that I'm in agreement with  @Largou walid  (and I'm in agreement that how one maintains consistency between environments/sites is an architectural decision, a decision where one considers operational policies and procedures and service levels.....and for these reasons "no one size fits all" 

    As far as what I advise clients, and have done so for over 20 years with WebSphere App Server, it's 

    -  perform all administration using scripting, this because "point and click  a GUI (the adminconsole) isn't repeatable, and without repeatable administration you can't meet HA and DR service levels 

    • after running the scripts in one environment, and validating the changes,
    • run the scripts in the next environment

    This is what I did when I worked for a living, before joining IBM, and it's how I met the availability SLAs mandated by the business units I supported 

    That said, with care one can employ replication, but you have to be careful, only replicating after validating the changes in the primary environment, in order to avoid the scenario I cited Additionally which one can change hostnames using wsadmin https://www.ibm.com/docs/en/was-nd/8.5.5?topic=servers-changing-node-host-names  when using replication, the IP addresses in each environment will be different, and this can cause issues with transaction recovery https://www.ibm.com/docs/en/was-nd/8.5.5?topic=recovery-configuring-manual-peer-transaction-service as the transaction manager in WAS has some IP Address dependencies



    ------------------------------
    Tom Alcott
    Senior Technical Staff Member
    IBM
    ------------------------------