Re: [sqlite] Windows I/O (was: Initial read speed greater than subsequent)

Experimenting with Data Warehouse - which should really be run on a more "mainstream" DB. Sqlite was supposed to be just for piloting and testing - but it's such an incredible little database engine - it's hard to let it go - so I try big things on it just for kicks - delaying the inevitable. It

Re: [sqlite] Initial read speed greater than subsequent

Thanks. Here is something that seems to work - I open a Sqlite "scratchpad" DB into which I select subsets from master tables - which are the problem because they are large. Initially I put each master table in its own DB and attached them all in the beginning of the script. The first read was

Re: [sqlite] Windows I/O (was: Initial read speed greater than subsequent)

I know the newer versions of Windows are fantastically bloated (and slower every version), but what are you running that uses more than 16 GB of committed memory? > Thanks. More RAM would clearly be helpful - but first I need a bigger > machine that can take it. For some reason - the "home"

Re: [sqlite] Windows I/O (was: Initial read speed greater than subsequent)

Thanks. More RAM would clearly be helpful - but first I need a bigger machine that can take it. For some reason - the "home" line of PC is typically capped at 16GB or so. I'll Need more of a workstation to go higher and experiment with the settings you suggested. On Fri, Jul 13, 2012 at 8:35 PM,

[sqlite] Windows I/O (was: Initial read speed greater than subsequent)

Windows is really atrociously bad at I/O. Windows has the same basic model of how to perform I/O as a 6 year-old. Scratch that, the six year old could probably understand I/O better than whoever wrote/designed the crap in Windows that passes for I/O routines. Anyway, make sure that you have

Re: [sqlite] Initial read speed greater than subsequent

On 14 Jul 2012, at 3:07am, Udi Karni wrote: > Thanks. I understand. I tried to set PRAGMA CACHE_SIZE=0; in the hope of > not cache anything and forcing Sqlite to always go to disk - but that > didn't help. > > I see some reads on pagefile.sys - but both the DB and the

Re: [sqlite] Initial read speed greater than subsequent

Thanks. I understand. I tried to set PRAGMA CACHE_SIZE=0; in the hope of not cache anything and forcing Sqlite to always go to disk - but that didn't help. I see some reads on pagefile.sys - but both the DB and the pagefile are on SSD - so you would think it shouldn't be too costly to read one or

Re: [sqlite] Initial read speed greater than subsequent

On 13/07/2012 5:37 PM, Udi Karni wrote: Hello, Running on Windows 7 - I am noticing that tables in :memory: DBs are read (SELECTED) at a constant rate. However - conventional DBs on disk - even on SSD - are read fast the first time, and much slower subsequently. Closing and reopening a DB for

[sqlite] Initial read speed greater than subsequent

Hello, Running on Windows 7 - I am noticing that tables in :memory: DBs are read (SELECTED) at a constant rate. However - conventional DBs on disk - even on SSD - are read fast the first time, and much slower subsequently. Closing and reopening a DB for every SQL statement seems to cure this -

[sqlite] leap seconds

The SQLite3 date & time functions are designed assuming > […] that every day is exactly 86400 seconds in duration. Before I start implementing TAI (or GPS time) to/from UTC translator plugin, has anyone already done this? Why? In a device that logs data with sub-second resolution, in my case a

Re: [sqlite] Native API for C# (C sharp) and Visual Studio

On 13 Jul 2012, at 3:11pm, joe.fis...@tanguaylab.com wrote: > We need to move a MS Access database to something more portable. Is the > “http://system.data.sqlite.org” the way to go for getting SQLite API access > from the Microsoft stuff (Visual Studio, other)? >

[sqlite] Native API for C# (C sharp) and Visual Studio

We need to move a MS Access database to something more portable. Is the “http://system.data.sqlite.org” the way to go for getting SQLite API access from the Microsoft stuff (Visual Studio, other)? The programmer needs to use C# (c sharpe) for his coding that interacts with the database. Is it a

Re: [sqlite] Correction needed in shell.c of 3071300

On Thu, Jul 12, 2012 at 10:25 AM, Roy Soltoff wrote: > Folks, in the latest linux amalgamation download, I needed to add braces > surrounding some of the code in the find_home_dir function to be able to > ccompile without an error. The added braces are identified by the

[sqlite] FTS: Phrase queries

Ever since I started using FTS extensively, I frequently ran into this limitation: ** TODO: Strangely, it is not possible to associate a column specifier ** with a quoted phrase, only with a single token. Not sure if this was ** an implementation artifact or an intentional decision when

[sqlite] Correction needed in shell.c of 3071300

Folks, in the latest linux amalgamation download, I needed to add braces surrounding some of the code in the find_home_dir function to be able to ccompile without an error. The added braces are identified by the arrows. The struct declaration cannot occur after a statement; thus the need for the

Re: [sqlite] Improving access speed to fetch indexed column.

The patch doesn't attached, please apply link to the patch. 2012/7/12 Filip Navara > This roughly resembles an issue I witnessed on our databases about > year ago (thread "Improving the query optimizer" on this mailing > list). SQLite doesn't use covering index for