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

Question

Question

High Speed Export Options for LaserFiche to File System

Laserfiche SDK Administration Integrations How To
Updated August 15, 2019
asked on August 14, 2019

As part of a merger/acquisition, we need to copy/export out 2+ TB of documents from Laserfiche to the file system to subsequently send them to another business partner(who will initially just load into a file share…they don’t plan to use Laserfiche).

One of our developers created a .NET program to export out, but the performance is pretty low (~ I GB an hour).   Any suggestions on high speed export of documents out of Laserfiche and onto the file system (documents and metadata) ?  We are on an older version (8.3).

0 0

Replies

replied on August 14, 2019

Do you know where the bottleneck is? It could be in code, CPU, bandwidth, or disk I/O.

I've only done this for massive imports, but I've written processes that use Parallel.ForEach to speed things up quite a bit. Then I ran the same process on multiple dedicated workstations, up to the point where it started to impact the performance of the Laserfiche server, and then I backed it off a little.

LFS can handle quite a bit, so I doubt it's the limiting factor. If you're converting the documents to PDFs, then there's overhead to doing that, plus other factors mentioned above.

2 0
replied on August 15, 2019

Thank you Devin for the response.  We are not sure yet where the bottleneck is, but are still looking.  All processes are running in same network segment.   There is no document conversion taking place.  We'll try the Parallel.ForEach.  I wasn't sure if there was an Admin function to do a high speed export or not.  I didn't see one but wanted to ask here if anyone knew.  Thanks again.

0 0
replied on August 15, 2019

Any time that I'm doing some kind of bulk process in Laserfiche, it's a two stage process. First, I'll dump a list of all of the entry ids that need to be acted on to a database table. That table will usually have the following columns:

  • Entry ID
  • IsMetadataDone
  • IsImageDone
  • IsError
  • ErrorText

 

I'll grab a set of rows from that table, and start walking through them, setting the fields appropriately as I go along. If there's a problem somewhere, it's pretty easy to restart where I left off.

Feel free to ask any clarifying questions.

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

Sign in to reply to this post.

Details
Related Posts
Loading ...