Replies: 3 comments 1 reply
-
You can store terabytes of data in PostgreSQL if your server is big enough. This shouldn't affect the performance. We could wonder why the files are stored as uuencoded strings and not as binary, which increases the volume by around one third. But that’s how things work for the moment. The more important question is if it makes more sense to store the documents inside SQL-Ledger or externally. In most of the cases we and our customers store the documents externally. This means we create codes with a URL shortener and add them to the transactions. Like that it is possible to make some funny stuff, like add a first entry that links to a PDF datasheet on SharePoint, a second one that links to the web shop of the supplier, and a third one that links to the warranty portal of the manufacturer. If on the other hand you use the reference documents for business records, you may have to consider the legal requirements. In our country we have the obligation to preserve the records for 10 years. In this case you’re better off if everything is stored in one place. If you have a tax audit, you just need a running SQL-Ledger to make all your documents available. There are still some technical questions. To reduce the amount of data we definitely need deduplication. And then you probably don’t want do have all the «And now a word from our sponsor»: I don’t want to forget to mention that most of the development of version 42 was sponsored by the company of @bentf4y. 👏 |
Beta Was this translation helpful? Give feedback.
-
I like how SL stores images. They are in Is there a specific reason for using different store systems for images and documents in SL? Also, I would like to thank @bentf4y sincerely for their kind support. |
Beta Was this translation helpful? Give feedback.
-
In case of a database corruption you need a database backup. In case of a file system corruption you need a file backup. This means you need two different systems to backup the whole installation.
I can't answer this question because it's not my invention. I would probably have solved it differently. |
Beta Was this translation helpful? Give feedback.
-
I have never used SL for uploading documents to the database.
Seeing a lot of nice improvements for reference documents in the last version (3.2.12.42), I wanted to give it a go and uploaded some documents. I see that they are stored in
archivedata
table. Each document comprises many many text lines inarchivedata
.I need some insights for the performance of database when uploaded documents reach hundreds. Is it a viable way of storing documents?
Please advise.
Beta Was this translation helpful? Give feedback.
All reactions