Need some TIPs: Managing Databases with many Documents

HCL Domino Domino Forum
1

I launch this topic in order to ear other opinions. I loved for you to participate and help me in a brainstorming to manage a situation

How can I manage a Database with more than 300.000 documents and 20 view?

Note: 20 of 300.000 documents create huge indexs. The Database has already more than 20GB.

Best Reagards

Antero Santos

2

Subject: Need some TIPs: Managing Databases with many Documents

I have a database with 40.000 documents and taked more then 4 GB disk space.Problem is the database get updated automatically everyday from our As/400.

In my current server (Xeon 2.8, 1GB RAM and RAID 5), users have take along time when they open it only in the morning, rather than that is acceptable.

I’m planning to create archive for the database.

If you have unused documents, consider create archive database.

3

Subject: Need some TIPs: Managing Databases with many Documents

Well data is data. I guess the more pivotal topic is what is the database doing? Meaning do you have a lot of computed column values in the views? Do you have many @dblookups on views based on a key? What is the database growth? How often are the document modified.

I had a client whose had data in a lotus notes database. There were over 1.3 million documents in the database and I was able to get all the documents quickly. What I did was to create a search engine functionality to search the database via full text searching.

But there are many ways to attack large databases.

HTH – Cheers

4

Subject: RE: Need some TIPs: Managing Databases with many Documents

This is a document management database. The column values are the values of the fields, there are no computed values in the columns. It uses @DBLookup inside the documents. The database growth is of about 300 documents per day, and the documents, once created and numbered, no longer suffer modifications by the users, but Agents can access and modify them. The idea of a search engine with full Text is very good. We already created one and it’s like you said. Searching documents using a search engine is the fastest way, nothing in comparison with the views, which are to slow.Never then less we need to work with the views, which mean that we have to find some way to improve performance.

Unfortunately due to client internal reasons, we can’t archive.

I would like to change more ideas about the search engine to improve the search. What is the size of the database of your client.

Thanks

Antero Santos