[web2py] Re: Web2py and support for more than a million of records

2015-07-13 Thread villas
Hi Alfonso You mention 'sqlgrids', but if using SQLFORM.grid do not forget to turn off the search widget. searchable = False. I think the grid should otherwise be fast because I think it paginates the data with limitby. Also be aware of unintended recursive selects. e.g. I've noticed in the

[web2py] Re: Web2py and support for more than a million of records

2015-07-08 Thread Leonel Câmara
It may be as the DAL stores every row in RAM and you're not using limitby to paginate the results. In the very near future you will be able to use an iterselect instead of select which fetches one row at a time to avoid this, it could solve your memory problems. Check this topic: https://groups

Re: [web2py] Re: Web2py and support for more than a million of records

2015-07-08 Thread Alfonso de la Guarda
Hi, I have notice this: When i use this data = db(db.subjects_sources.media_group==1).select( db.subjects_sources.id, db.subjects_sources.name, db.subjects_sources.media_url, ) My computer just hangs (8 GB RAM), growing slowly the RAM, media_group is indexed. How

[web2py] Re: Web2py and support for more than a million of records

2015-07-08 Thread Leonel Câmara
If your table is really big it's very important to pick which columns you want in your selects. In particular, if you're using the grid make sure you use the fields keyword argument and only pick the fields you really need. If you usually sort/search this grid using some of the fields, make sure