As any other archiving solution, it is very well suited for environments with a large footprint of unstructured data (CIFS / NFS shares for user data) with a large amount of unused/old files and a need to keep those unused files for long term. In our scenario, due to some legal and contractual constraints we need to keep these files for 15 years. Archiving is a good choice to move the unused files to a cheaper storage tier, both on-prem or cloud.
Oracle Coherence can be used to solve latency problems by caching data near the application tier. In-memory performance helps to reduce data contention, thus improving application response time. Oracle Coherence does a great job at scaling linearly and can do so dynamically. Oracle Coherence can replicate data so it can be part of a disaster recovery solution. One thing that is often overlooked with elastic caches is their ability to analyze data in memory, leveraging the processing power of the data grid. This is something Oracle Coherence does exceptionally well. Oracle Coherence also provided event handling capabilities to allow applications to respond to events triggered by transactions.
Oracle Coherence support team is responsive and knowledgeable. We contacted them to ask a couple of design questions about how we were setting up Oracle Coherence based on how we used IBM's Datapower Extreme Scale. They were able to guide us so that we got the design correct the first time and didn't have to go back and re-architect our design later.
We have used Veritas Enterprise Vault in the past, and besides its being a well-known player on the data archiving market, their tool is far more complex to implement, to manage and to keep working. Komprise is very robust and also very easy to implement, as most part of the job is done on Komprise side. The management console is delivered through a public URL as a SaaS platform. You only need to deploy a few VMs for scan/archiving/user access, which they call "Observer VMs." Komprise also doesn't uses Stub files, which is a poor implementation adopted by the competitor for file access. We had a lot of issues in the past with stub files. Komprise has implemented 'bread crumbs', which are CIFS symlinks to the files on the Observer. It is a very good implementation and it works really well.
DataSynapse GridServer has the same cache functionality connected with a grid distributed application deployment platform. It provides all the necessary tools to configure and manage these applications. On the otherhand Oracle Coherence has more a flexible configuration and better performance. Also the Oracle configuration and monitoring tools are more convenient and informative than the DataSynapse ones.
In my current project, it is yet to be decided as we are on the way to make it live for users. But I am sure it will be an added value to obtain timely calculations.
