> People who understand SQL can employ the sqlite3 command-line shell to analyze large datasets.
And a bit further down:
> SQLite database is limited in size to 140 terabytes [...] if you are contemplating databases of this magnitude [use something else]
Yeah no. "Large datasets" here means a few megabytes. I figured that out the hard way:
I had a database of about 70 megabytes and ran a query with "COUNT(a)" and "GROUP BY b" on it. This makes it write multiple gigabytes to /tmp until it goes "out of disk space" (yeah /tmp on my ssd isn't large).
I heard nothing but awesome and success stories about SQLite until a few weeks ago when this fiasco happened. I still like SQLite for its simplicity and last week I used it again for another project, but analyzing "large" datasets? Maybe with a simple SELECT WHERE query, but don't try anything more fancy than that when you have 100k+ rows.
And a bit further down:
> SQLite database is limited in size to 140 terabytes [...] if you are contemplating databases of this magnitude [use something else]
Yeah no. "Large datasets" here means a few megabytes. I figured that out the hard way:
I had a database of about 70 megabytes and ran a query with "COUNT(a)" and "GROUP BY b" on it. This makes it write multiple gigabytes to /tmp until it goes "out of disk space" (yeah /tmp on my ssd isn't large).
I heard nothing but awesome and success stories about SQLite until a few weeks ago when this fiasco happened. I still like SQLite for its simplicity and last week I used it again for another project, but analyzing "large" datasets? Maybe with a simple SELECT WHERE query, but don't try anything more fancy than that when you have 100k+ rows.