High Availability, Database Mirroring plans for servers

Respuesta propuesta High Availability, Database Mirroring plans for servers

  • 30 ธันวาคม 2554 17:07
     
     

    Hi All

    I have a question, I have two physical servers, in the first one I'll install two virtual machines in order to have two DB Servers and the other I'll use as DB Server support. The idea is to have three servers, this is the role for each one

    VM1.- it will be used for sandbox, development and testing enevironments

    VM2.- it will be used for staging and production environment

    DBServerSupport it will be used when any of servers mentioned before breakdowns.

    I want to implement the mirroring in High Availability mode in VMs, It means that VM1 will be the principal and DBServerSupport will be Mirror, the same situation for VM2 (I'll use other server as witness in both situations).

    It is possible to have that environment?

     

     

     

     

     

     


    Jesus Angel

ตอบทั้งหมด

  • 31 ธันวาคม 2554 12:25
     
     คำตอบที่เสนอ
    This seems to be fine but please make sure DBServersupport server should be capable of  handling load when failover happens from both VMs 
    http://uk.linkedin.com/in/ramjaddu
    • เสนอเป็นคำตอบโดย PrinceLuciferMVP 1 มกราคม 2555 14:57
    •  
  • 1 มกราคม 2555 6:55
     
     
    Do you want VM1 to be a Principal and  DBServerSupport and VM2 as a mirror? In that case I think you need Log Shipping....
    Best Regards, Uri Dimant SQL Server MVP http://dimantdatabasesolutions.blogspot.com/ http://sqlblog.com/blogs/uri_dimant/
  • 4 มกราคม 2555 16:18
     
     

    Hi Uri Dimant

    The idea is that both VM1 and VM1 play the principal role, and DBServerSupport plays the Mirror role, my doubt is if it is possible?, I mean to establish high availability in two different environments where the mirror is the same server for both


    Jesus Angel
  • 5 มกราคม 2555 15:13
     
     

    I believe you could configure it that way, but one really has to ask why you would?  Meaning, hardware is cheap when compared to the human capital, VMs are SLOW and you retain a single point of failure (the VM host). 

     

    Consider this, your Support server would be alone if the VM host took a hit HENCE, in a high availability mode, your database(s) would be off line as the witness and mirror are no longer present (quorum; you always need two instances to see each other and you will not have that with a vm as principal (or mirror), the other vm as witness when the VM host goes off line).

  • 6 มกราคม 2555 16:00
     
     

    Hi

    I agree with you about the point of failure, but by the time, I would like to implement this environment because of current resources, and it will allow me to request for more resources in near future, it based on a document where it would includes all details that affect to High Availability environment such as the point of failure and the hardware features.

     

     


    Jesus Angel