Adding redundancy for data availability
Use the following information to understand basic concepts and scenarios that you can use while adding redundancy to your system.
The following video (1:26) illustrates the need for applying redundancy in your environment.
Data flow within the TrueSight IT Data Analytics framework
Data moves through TrueSight IT Data Analytics via different channels, as depicted in the following figure.
The channels depicted in the preceding figure are explained as follows:
|Collection Agent and Collection Station
The Collection Agent collects data and sends it to the Collection Station.
The Collection Station receives data and forwards it to the Indexers for indexing.
|Collection Station and Indexer
|The Indexer receives data from the Collection Station and indexes the data to make it ready for search.
|Search and Indexer
The Search components contact the Indexers for data as per the search requests made by various users.
TrueSight IT Data Analytics provides you a mechanism of collecting and searching data. In a multiple-server deployment, you can have the Collection Station, Indexer, and Search components installed on various nodes. If one of these nodes goes down, you can start losing data and thereby valuable knowledge that might be crucial to your business.
Loss of data can occur at different stages of data flow within the TrueSight IT Data Analytics framework. Depending on which node goes down, you can experience data loss at the data collection stage, indexing stage, or search stage. For instance, if the Collection Station goes down, data collected by the Collection Agents will not reach the Indexer and therefore will not be available for search. If the Indexer goes down, the data collected will not be indexed and therefore will not be available for search. Similarly, if the Search node goes down, the data collected and indexed will not be searchable.
This problem can be solved by adding redundancy to your system. Redundancy means that if one node goes down, another node in your system takes up the job of the first node. In this way, data continues to be collected, indexed, and searched.
Need for redundancy
The need for redundancy depends on your business needs. If you want to increase data continuity and availability you need to apply redundancy. Also, if the data you are collecting is critical, you must apply redundancy. Redundancy can only be enabled if you are operating in a multiple-server deployment. This is because redundancy is only applicable when you are operating in an environment with multiple Collection Stations and Indexers.
Enabling redundancy has a cost in terms of hardware resources required. However, in the long run redundancy can save the cost associated with losing data. Depending on the data availability needs of your business, you need to manage the trade-off between the benefits of data availability and the cost of hardware resources required (for enabling redundancy). For more information, see Sizing drivers and their impact.
Supported redundancy scenarios
The following redundancy scenarios are supported: