Accepted Solutions
Kevin -
- It should be noted that Data Insight is not cluster aware.
- We do not support installing it within a cluster.
- Data Insight is not able to scan or monitor itself for changes.
I agree that using the Veritas Infoscale product suite is likely your best opportunity for success.
That said there are premises one would need to familiarize themselves with to envision the configuration you would require to have a successful Disaster Recovery (DR) within a Data Insight environment.
For example the databases have to be synchronized to the configuration and all updates must succeed in multiple places. In a clustered environment it could be fine to use shared storage and have different nodes active when administering the application but then each node would have to be able to change to a state of controlling the environment. The cluster must actually be the node as it is imperative that the Management Server (MS) can resolve each node name to a configuration file including its' own. Very likely the understanding required to visualize the environment and each node's contribution to the application's operation is outside of the scope of a blog article.
Since you have only asked about the MS being clustered and not the databases themselves we have lost the failover or DR protection the question indicates is desirable for anything but the simplest single tier environment. Take for example this simple diagram:
Our considerations would have to be:
- IdenticalDNS entries for all nodes and sites since the worker nodes are registered to a Name and/or IP
- Replication technologies or methods to ensure the synchronization of Database copies
- Product, platform, configuration and database schema versions must match
- Automated application start (Services on preregistered worker nodes)
There are likely considerations I have forgotten to include however research / testing should be used to confirm operation prior to inclusion in any production environment. My suggestion would be to initiate a Veritas Support case to have guidance for the inevitable questions and testing scenarios to ensure you have the most up-to-date advice for your planned implementation.
Hopefully this information is of use to you in making your decision.
Rod
Kevin -
- It should be noted that Data Insight is not cluster aware.
- We do not support installing it within a cluster.
- Data Insight is not able to scan or monitor itself for changes.
I agree that using the Veritas Infoscale product suite is likely your best opportunity for success.
That said there are premises one would need to familiarize themselves with to envision the configuration you would require to have a successful Disaster Recovery (DR) within a Data Insight environment.
For example the databases have to be synchronized to the configuration and all updates must succeed in multiple places. In a clustered environment it could be fine to use shared storage and have different nodes active when administering the application but then each node would have to be able to change to a state of controlling the environment. The cluster must actually be the node as it is imperative that the Management Server (MS) can resolve each node name to a configuration file including its' own. Very likely the understanding required to visualize the environment and each node's contribution to the application's operation is outside of the scope of a blog article.
Since you have only asked about the MS being clustered and not the databases themselves we have lost the failover or DR protection the question indicates is desirable for anything but the simplest single tier environment. Take for example this simple diagram:
Our considerations would have to be:
- IdenticalDNS entries for all nodes and sites since the worker nodes are registered to a Name and/or IP
- Replication technologies or methods to ensure the synchronization of Database copies
- Product, platform, configuration and database schema versions must match
- Automated application start (Services on preregistered worker nodes)
There are likely considerations I have forgotten to include however research / testing should be used to confirm operation prior to inclusion in any production environment. My suggestion would be to initiate a Veritas Support case to have guidance for the inevitable questions and testing scenarios to ensure you have the most up-to-date advice for your planned implementation.
Hopefully this information is of use to you in making your decision.
Rod
