Understanding the Hypertrieve database engine

The Hypertrieve database engine is a highly efficient, object-oriented database that is developed by BlueCielo ECM Solutions and embedded into Meridian Enterprise. It is optimized for use with Meridian and provided so that Meridian customers can implement engineering content management system without the need for additional software.

When you create a new Hypertrieve vault, you are given a choice of multiple Hypertrieve database engines. The first generation Hypertrieve database engine is no longer available. The database engines with names that end with the number 3 are second generation engines that have been extensively optimized compared to the first generation, particularly for 64-bit processors. The database engines with names that end with the number 5 are third generation engines that have been optimized for scalability compared to the previous generations. This optimization is primarily through the use of multithreading and for use with the Use HyperCache option described in Configuring HyperCache.

Before migrating an existing vault to Hypertrieve 5, first consider the other optimization techniques described in Optimizing performance , especially the HyperCache option, which usually boosts system performance significantly.

To determine if Hypertrieve 5 can provide additional benefits over Hypertrieve 3:

1. Run the Windows Performance Monitor for several working days as described in Configuring the Windows Performance Monitor. Select all performance counters for the process AMEDMW.exe and all counters in the HyperTrieve group.

2. Review the collected data for the following:

   • The utilization of the single processor/core used by Meridian rises to close to full utilization during peak usage hours and remains there for one or more hours.

   Note    The counter represents the total capacity of all CPUs/cores. For example, on a 4 processor computer, full utilization of 1 CPU/core is shown as 25%.

   • The Engine queue length counter in the AutoManager Hypertrieve Statistics group often rises above 1.

If performance monitoring reveals that both of the preceding conditions occur frequently, then it is likely that Hypertrieve 5 will benefit your system. Hypertrieve 5 will make much better use of the available hardware by spreading the system load over all available processors/cores. Meridian will then provide faster and more consistent response times to its users. With other Hypertrieve versions, the user response times are highly correlated with the load on the single processor/core that is used.

Notes

• To migrate an existing vault from one generation of database engine to another, you must restore a backup of the source vault after first setting the destination database engine name in the DLLName setting in the vault's DataStore.ini file. The general procedure for restoring a vault is described in Moving a Hypertrieve vault.

Hypertrieve requires very little system administration and, in fact, has no database management user interface. The data files for a vault stored in Hypertrieve reside in the location specified when the vault was created, C:\BC-Meridian Vaults by default. Each of the Hypertrieve files are named <VaultName>.* where each file extension is described in the following table.