db-derby-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From BONNEFILLE Guilhem <guilhem.bonnefi...@c-s.fr>
Subject Reducing disk footprint with compression (gzip?)
Date Thu, 14 Apr 2011 16:58:10 GMT

I'm completly new to this list, so I hope you will be patient with me.

We are using Derby in a sort of bi-dimensionnal database. Database's  
schema is the following:
* a blob (around 100 bytes)
* a timestamp
* an identifier.
The use case is:
* a single continuous input source,
* many parallel requests, time oriented for a given list of  
identifiers, to extract chunk of database.

Currently, performances are quite good.

The matter occurs with disk footprint: disk usage is too heavy.

One solution to reduce this is to modify the schema (a little bit more  
complex than what I described above).

An other solution is to compress (gzip) old part of history as we  
never need to update them.

Is it possible to do this with Derby?
Do you have any other tips when faced to this use case?
Any other (embedded) database that best matches our needs?

Thanks for your answer.

This message was sent using IMP, the Internet Messaging Program.

View raw message