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

Question

Question

PsiGen integration

asked on December 5, 2016 Show version history

We have a customer that uses the PsiGen integration and they're running into some issues with metadata being incorrect when it a document comes into Laserfiche. There is a workflow running on those documents as soon as they're sent to Laserfiche, but the workflow's first activity is to retrieve field values, and we can see that the values it retrieves are incorrect, leading us to believe workflow isn't messing up the field values. Additionally, some documents' metadata is so messed up it doesn't even fulfill the workflow's starting rules, further convincing us it's not a Laserfiche issue. This only happens to a few documents in a batch - less than 5%.

Has anyone had troubles similar to this? The customer hasn't been able to have their PsiGen techs take a close look yet, and it feels like we've done all we can on the Laserfiche side to try to pinpoint where the issue is. Anyone have any tips?

Thanks!

0 0

Replies

replied on December 5, 2016

It's hard to guess without actually looking at the data, but the most common problem with integrations that causes this behavior is that they create the document and set fields as 2 separate events. So Workflow can get the notification on the document creation and reads its fields before they're set. It's a timing issue depending on the load on LFS/SQL/WF, which is why you're not seeing it on all documents.

You could look into modifying your starting rules to look for the Entry Changed event instead. You'll have to most likely tweak the workflow as well to make sure it only starts once.

0 0
replied on December 5, 2016

PsiGen has a stage where you can view the metadata and perform QA on it before the batch is "migrated", similar to Quick Fields. What does the metadata for those document look like there?

One thing you can do to troubleshoot the issue is to have to Migration stages, and have the second one export the metadata as XML. Then you can cross-reference the documents with corrupted metadata by looking for them in the XML file.

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

Sign in to reply to this post.