High Availability and HP TRIM

By Jason Boswell


HP TRIM is deployed by many organizations around the world as a mission critical system, as such HP is often asked about the High Availability of TRIM. There is no single answer to this question. High Availability is more of a strategy than a definable design, and each organization will find that they have a unique set of requirements, and constraints which will help shape the nature of a highly available HP TRIM solution. This article will attempt to help an organization understand the nature of the components of HP TRIM and how each relates to High Availability, so that those responsible for implementing the High Availability strategy are able to make informed decisions as to the best approach.


Outages (planned and unplanned)


A couple of distinctions need to be well understood from the outset: The first being the difference between Planned and Unplanned outages. All mission critical applications require maintenance at some time, and HP TRIM is no different. These scheduled updates, patches or maintenance are considered to be Planned Downtime, and should be factored into High Availability requirements. Another closely related term to High Availability is Recovery Time, which is the average time to recover from an unplanned outage, or disaster. It should also be noted that in some cases recovery time may be infinite, and acceptable versus the cost of providing a scenario that diminishes the risk of this unrecoverable situation.


Outside of any technical issues relating to the provision of high availability, an organization must consider the level of risk or level of service that they are willing to bear. This understanding of the risk/reward is just as important as any discussion about technology. Organizations must understand that given infinite budgets and infinite resources, there is still the potential for unplanned outages of any system. It is in this light that HP structures this article. 


HP TRIM Architectural Components


HP TRIM is made up of a number of architectural components. Each of these has its own unique properties which need to be considered in a High Availability strategy. It is also necessary to understand which components provide a higher value in terms of criticality. HP TRIM is broken up into three distinct tiers. The client tier, the HP TRIM Context service tier and the Data tier, as illustrated below.


HP TRIM tiers


HP TRIM is designed such that Clients never have direct access to the Data tier. This provides a measure of security by design, but also allows for greater scalability than a traditional client/server model. Thus it can be said that a Client communicates with the HP TRIM tier, and the HP TRIM tier communicates with the Data tier on the Clients behalf. For High Availability discussions this is a key concept for understanding the communication path of the user through to the servers and data.


Client Tier


The Client tier is made up of the supplied HP TRIM desktop or deployed applications, the web interfaces and the SDK or custom applications. HP TRIM client applications can be configured to automatically failover to different workgroup servers, and should not require too much in the way of High Availability planning. Obviously, a user that has a broken or failed ‘client’ will be unable to communicate with the HP TRIM tier, but providing something as simple as an alternative client (potentially web based) or a help desk that can easily deploy the desktop package, will provide a high level of recoverability, in a situation where incremental investment in redundancy provides very little extra availability.


Data Tier


The HP TRIM Data tier is made up of two distinct parts. The Database and the Document Stores. The database is the metadata store; all information and configuration data is stored in the database. This component is critical to the operation of the system.


HP TRIM utilizes traditional RDBMS systems, and therefore proven High Availability measures for these systems can be implemented. HP suggests the organization understand the recommendations provided by the RDBMS vendor.


Document Stores are a critical component of the HP TRIM system, housing all the electronic documents & records that are committed to the system. HP TRIM Document Stores also support traditional storage methods. HP recommends the organization seek advice in regards to high availability of the store. Document Stores can utilize existing storage areas available via CIFS (e.g. HP StorageWorks NAS) or through HP's Integrated Archive Platform (IAP).


Components of the system that are also stored as a part of the Data Tier includes logs, audit trails, and the Document Content Index. This data is not imperative to the operation of the system. However, it is required to ensure the authenticity and integrity of the information contained in the system. For simplicity, these data components can be considered to be like a document store. Organizations should seek advice from their storage vendor regarding High Availability.


Service Tier


The HP TRIM Context Service Tier provides a variety of services to the Client tier. Each of these services provides different value to the system, and can be deployed on different servers in different geographic locations. This document will not discuss the pros and cons of where to deploy services, but it should be noted that geographic location and co-existence of these services needs to be considered in any High Availability planning. There is no preference for them to coexist or be centralized etc., but should be noted that these infrastructure decisions affect High Availability planning.


HP TRIM Workgroup Service – this service provides the connectivity between the client and the data tier and manages document caches. Each client is able to communicate with a primary and a secondary workgroup server, and will automatically failover between these two. There is no critical data stored as a part of a workgroup service; however, a document cache will affect network traffic if it is lost.


HP TRIM Event Processor – This service provides data transaction support for the system. As the name suggests it processes events relating to the system, including:



  • Mail notifications

  • Audit and billing logs

  • Background Records Management tasks (automatic part creation and scheduled triggers)

  • Indexing of words and content

  • Custom events

No critical information is stored as a part of a standard event processor.  However, that cannot be said for any HP TRIM Event Processor Custom events. The service while critical in terms of longer term accessibility of documents (being indexed and searchable) and ensuring Records management tasks and mail notifications are current, is not required for the system to provide availability to users.


Web Services/interfaces – HP TRIM provides numerous web interfaces as a part of the system. They can be considered as a client in terms of availability. However, they are primarily a service from a technology point of view.


Ice – is a customizable web server. This is an interface to the primary HP TRIM system. In itself it houses no critical information, and multiple versions or instances of Ice can point to the same HP TRIM environment. There are no specific High Availability requirements, and organizations need to consider the ease of recovery when choosing to deploy multiple redundant Ice instances.


SharePoint Integration – HP’s SharePoint integration is a series of webparts to communicate with HP TRIM via SharePoint. All recommendations for High Availability for SharePoint should be garnered from the reseller of SharePoint or Microsoft.


The question of High Availability is one that is difficult to answer for all. There are a number of factors that need to be considered and depending on the makeup of the organization, each will have a different ‘value’ in the makeup of the overall strategy. Resources such as committed budget, available staff and support, criticality of the system and the consequences of outages, as well as the situations themselves can all play a role in the need for high availability. However, it should be said that systems that need to withstand a catastrophic attack should not have the same level of High Availability to those systems that require users to provide answers to the public in a ‘timely’ fashion. It is this very concept that leads HP to state that High Availability is more of a strategy than a one-size-fits-all answer.


As the developer of HP TRIM, HP may be consulted to help you make a better informed decision as how best to implement a High Availability strategy.


 

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
About the Author


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation