 How does DISA reduce bandwidth requirement? |
 |
| This innovative technology recognizes that by placing objects closer to the places of usage in a distributed server environment, instead of a monolithic data-centric model. The DISA allows the client to decide whether to send the request over a LAN or a WAN to access the data. This improves speeds, balances load, and reduces bandwidth manifold.
The DISA architecture also allows us react to temporal bottlenecks and infrastructure breakdowns, but taking advantage of the inherent redundancies in the network design.
|
|
| How scalable is the DataScan.Online architecture? |
|
| DataScan.Online is scalable and flexible. It can scale from your laptop to cover the complete enterprise. The system is designed to make use of all available resources optimally.
Within a server it scales up according to the available RAM and also CPUs. In an enterprise, it can be scaled to tackle either traffic increase or document content volume increase or both.
There is no limit to the number of servers that form the DISA backbone, or to the amount of document content that can be stored under this system or the number of storage units that it can access.
|
|
| Why is a specialized database used to store the data instead of the popular RDBMS platforms? |
|
| The high-performance ImageBase database was designed to specifically address the peculiar needs of a CMS system, which are not handled well by the RDBMS platform. These needs are:
· Ability to manage Binary Large Objects (BLOBS) effectively within the system.
· Ability to treat all keywords, and combination of keywords as database indices.
· Ability to handle multi-value data type or “sets”.
· Ability to handle a heterogeneous storage environment spread over wide geographical locations.
|
|
| What are the similarities and dissimilarities of the ImageBase storage structure from a traditional RDBMS structure? |
|
| The documents/objects in the ImageBase is stored logically as rows of fields as in an RDBMS system. However, whereas each field in an RDBMS contains one value for each row, the ImageBase structure allows more multiple values. For example, the personal file of an employee in the HR Dept. can be tagged with Qualification, Languages Spoken, and Skill Sets, all of which will have a set of values instead of a unique value. RDBMS workarounds are possible, though not as efficiently and seamlessly as the ImageBase.
Each table in an RDBMS is normally associated for one or more index files, to facilitate faster access. However, all implementation of RDBMS puts limitations on the number of index files per table, along with the composition of the fields in the key. These occupy extra storage space.
The unique design of the ImageBase removes all limitations on the number of index files, or the composition of the index. It also does not take extra space, and does not create any extra files to facilitate this large number of logical index accesses.
|
|
| How is the DISA strategy different from clustered database strategies of other databases? |
|
| Access of data from clustered servers is routed through a server. This means that the bandwidth between the clients and this server is used frequently. However, in the DISA architecture, a thin router layer is maintained at the client. This allows routing to take place from the client directly to the relevant server. No WAN connection is heavily used, while access to local data from local server requires no WAN bandwidth. |
|
| DISA and mobile access? |
|
| Yes. We can combine the client, database servers and member server in one laptop running Microsoft’s Personal WEB server. When in office, the application server in the laptop can be working as a member server, managing personal data on the local hard disk of the laptop. However, when you are travelling, you can then start the application server as an authentication server and access local data. |
|
| How does DISA use the infrastructure redundancies to improve reliability? |
|
| The infrastructure elements in the system are communication lines, application servers, database servers and storage units. If a particular communication channel is down, then the data can be routed through the hardware switches. However, if a member server is down, then the ImageBases allocated to it can be temporarily re-assigned to another member server in the network. If a database server or the storage unit is down, then we cannot recover unless disk replication services are used. In case the WAN goes down completely, then the local member servers can be restarted as authentication servers, allowing local data access to continue. SO even in this extreme situation, the system allows the operations to continue, even in a slightly reduced manner. |
|
| Can we have document content for two domains stored within the same database server? |
|
| Yes. The design of the ImageBase location strategy is totally divorced from the domain management. The domain related application servers conduct information and decision making. |
|
| How are domains managed? |
|
| Each domain can be assigned to more than one application server. Each application server can access document content belonging to multiple domains.
The authentication server facilitates exchange of information between itself and other member servers, allowing for domain knowledge are distributed to all the servers accessing that particular domain.
All management functions are controlled from the authentication server.
|
|
| Is the DISA scalable? |
|
| The DISA strategy allows the system to be built up block by block. As explained above there are many infrastructure elements, whose sizing affects the overall performance of the system.
The DISA allows your organization to start with current data, and then grow (or shrink) in sophistication, power and expanse as the needs change over time. Sometimes, bandwidths needs to be changes, sometimes storage units need to be added. DISA allows to built-up a large enterprise systems by linking the infrastructure resources in a most effective manner.
|
|
| How does DataScan.Online maintain history? |
|
| Most software when required to maintain a record of the older versions, would typically duplicate the document content. However, when we are dealing with documents of pages of information, and the size of the page objects becoming larger, this method is typically very expensive on storage.
DataScan.Online avoids this problem, by intelligently storing only the parts of the documents that have been affected by the version changes.
|
|
