On Thu, 23 Mar 2017 13:45:58 -0400 Stephen Chrzanowski <pontia...@gmail.com> wrote:
> Right up front, I'm fully aware about the How To Corrupt document > (Believe me, I've preached about network access in this forum), and > with the development tools I have in Windows vs the destination OS > and the purpose of the DB, I'm asking for other peoples experience on > remote developing a database. > > The SQLite editor of choice for me is SQLite Expert Pro (SEP). The > remote system is a Linux based OS. The databases job is to keep > track of jobs, hosts, last completed, priorities of the jobs, etc. > The Linux machine is going to be running a BASH script that runs in > an infinite loop, periodically poking the database to decide what to > run next based on a schedule. There will be frequent sleep periods > between SQL calls. > > While I'm developing the database, the infinite looping in the bash > script isn't going to exist. The script runs, does its thing (To > start, just ECHO what I want it to do), update the database on > successful completion, then check for the next job if any are > available. When the scripts are done running, I want to re-run a > query in the SEP to confirm what I've done in the BASH script did > what it was supposed to do. > > The question for the experienced multi-machine & multi-OS DB > designers, has anyone ever run into a problem where EXTREMELY LIGHT > WEIGHT use of the database causes corruption? What would be a > recommended way to setup the connections for a DEV-only arena where > the below paragraph describes? > > By EXTREMELY LIGHT WEIGHT use, I mean I *DO* guarantee that although > I have one permanent open file handle to the database via SEP, and > that Linux OS will only open a handle periodically while I'm writing > the script, multiple accesses of reading or writing to the DB at the > exact same time just will not happen. Once development stops, it'll > be just this one BASH script that will ever touch the database. I'll try this manner: a) write a file with the query in the server b) bash script pass it to sqlite3 command line interface and it writes output to other text file c) read the file with the result d) delete the files You can grow or complicate the steps as you need. For the network server code, repository has an example server http://www.sqlite.org/src/artifact/a2615049954cbb9c and timeline at http://www.sqlite.org/src/finfo?name=src/test_server.c --- --- Eduardo Morras <emorr...@yahoo.es> _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
