Repstor custodian – Take a look at our Provisioning Engine Scalability

Repstor custodian and our Provisioning Engine are designed from the ground up to support very large deployments, with minimal performance impact on SharePoint…






Repstor custodian and our Provisioning Engine are designed from the ground up to support very large deployments, with minimal performance impact on SharePoint. In other words, if your SharePoint environment can handle a million matters or cases, then so can Repstor custodian. Our browsing and search interfaces will continue to work flawlessly even under these tough conditions.

How do we do that?

Firstly, custodian data is cached with Redis. This helps in two areas:

  • Supports multiple custodian servers and allows their data to stay in sync – our servers are ‘stateless’
  • Speeds up the interface by preventing frequent and unnecessary SharePoint querying

Our Open Source Level 1 in-Memory cache also ensures that Redis itself does not need to be contacted more frequently than is necessary; after all, the fastest number of bytes to transfer over a network is 0!

Secondly, querying SharePoint lists can be a delicate operation when you need to remain scalable. We’ve recently written a lengthy article detailing a number of the CAML querying concepts we use to stay performant.

The Repstor custodian application connects to the Provisioning Engine entirely asynchronously, with queues. This allows us to deploy any number of custodian servers and any number of provisioning engine instances – and a queue will sit in the middle ensuring that there’s no backlog of data. The type of queue is configurable, but MSMQ, Azure Service Bus, and Azure Queues are some examples we support.

Repstor custodian and provisioning engine scalability
Repstor custodian & Provisioning Engine Scalability

In a Microsoft Azure environment, both custodian and the provisioning engine will scale out independently, depending on load. The queue will never contain a significant build up as long as SharePoint can keep up. In an on-premise environment, it’s easy for systems administrators to monitor the queue and respond accordingly, but even during peak times, SharePoint is given time to catch up with its provisioning tasks as we incrementally back-off when SharePoint reports that it’s under heavy load.

Repstor custodian and the Provisioning Engine will ensure that the bottleneck in any deployment will always be SharePoint.

To learn more about Repstor custodian visit our website at www.repstor.com/products-for-ecm-adoption/repstor-custodian.

Written by Jonathan Cardy, Product Architect, Repstor.