You are viewing limited content. For full access, please sign in.

Discussion

Discussion

Ideal Laserfiche Implementation

posted on April 12, 2017 Show version history

I am studying for the 2016 MCSE and noticed that Microsoft recommends running single-task servers. For instance, running a server with just IIS or Hyper-V, etc.

Does Laserfiche recommend running single-task or single-role servers? For example, running Full-Text and Search on one server or Forms Routing Server Service on one server with Forms site installed on an IIS Server.

This type of implementation I assume would assist you in monitoring server resources, minimize laserfiche impact as a whole, and establish a failover server. I assume this would only be implemented with enterprise clients utilizing RIO.

Do you have documentation on an ideal VM Server implementation?

 

 

0 0
replied on April 12, 2017

Whether or not you should segregate Laserfiche components to their own dedicated servers depends on several factors.

Primarily, it's a matter of what type of resource utilization you expect. We always have a dedicated web server (i.e. a single-role server) that hosts Web Access and Forms; in our experience, they tend to be quite resource-intensive, but can be reliably run on the same server. Workflow services go on their own server as well, and so does the LF Server service. With Full-Text Search, it's a matter of how much indexing and full-text searching is going on in the repository. For example if you are scanning and OCR'ing lots of documents with Quick Fields and users are then text-searching those documents, it might make sense to deploy it on its own server. We also have one installation where Import Agent does a lot of OCR during import, so we put it on its own server that has a dual-core, high clock-speed CPU. In that case, that server can be considered a single-task server, since it does just one thing.

The other factor you may want to consider is ease of server administration. Keeping components segregated allow you to perform maintenance on their servers without affecting the other services. For most organizations this isn't a huge issue since maintenance tends to be performed overnight or over weekends, but if it's a high-availability environment then even a simple system reboot may need to be scheduled in advance and planned so that it's impact is minimized.

 

1 0
You are not allowed to follow up in this post.

Sign in to reply to this post.