Re: [web2py] Scheduler - a second try

2013-11-12 Thread Johann Spies
Thanks for your answer. Apologies. I have only seen your email now. I will try tomorrow. I have to go in a few minutes. Regards Johann On 12 November 2013 10:57, Niphlod wrote: > are the tables created on the server (not just scheduler_task, but also > scheduler_worker) > I find hard to bel

Re: [web2py] Scheduler - a second try

2013-11-12 Thread Niphlod
are the tables created on the server (not just scheduler_task, but also scheduler_worker) I find hard to believe (or, at least, I was never able to verify) that on the first round with only one worker the worker itself can't successfully commit a query that basically doesn't do nothing (because

Re: [web2py] Scheduler - a second try

2013-11-11 Thread Johann Spies
Just one. I never had any success with the scheduler and am trying to get just one instance to work. I did include my configuration in the first email in this thread. Regards Johann On 11 November 2013 13:04, Niphlod wrote: > how many schedulers are running ? just one with the default config

Re: [web2py] Scheduler - a second try

2013-11-11 Thread Niphlod
how many schedulers are running ? just one with the default configurations ? On Monday, November 11, 2013 7:08:38 AM UTC+1, Johann Spies wrote: > > > On 9 November 2013 15:05, Niphlod > wrote: > >> >> @spies: sorry for the digressions, any luck figuring out what database >> backend you're using ?

Re: [web2py] Scheduler - a second try

2013-11-10 Thread Johann Spies
On 9 November 2013 15:05, Niphlod wrote: > > @spies: sorry for the digressions, any luck figuring out what database > backend you're using ? > Yes. I replied earlier in this thread: Postgresql. -- Because experiencing your loyal love is better than life itself, my lips will praise you. (Psal

Re: [web2py] Scheduler - a second try

2013-11-09 Thread Niphlod
On my not-so-limited experience with SQLite and multiprocessing, I found out that it's a real PITA even with WAL enabled. If you turn on DEBUG logging, the amount of lines logged is going to block operations just for logging sake, so I'd definitely not recommend SQLite as a logging backend for

Re: [web2py] Scheduler - a second try

2013-11-08 Thread Jonathan Lundell
On 8 Nov 2013, at 11:40 AM, Niphlod wrote: > of course the syslog-ng makes sense only for the scheduler (but, if you're > using uwsgi with multiple processes, syslog-ng is the only handler that works > without issues, for the exact same "limitation") > Logging into database (if not something l

Re: [web2py] Scheduler - a second try

2013-11-08 Thread Niphlod
of course the syslog-ng makes sense only for the scheduler (but, if you're using uwsgi with multiple processes, syslog-ng is the only handler that works without issues, for the exact same "limitation") Logging into database (if not something like mongodb & co.) seems a little bit daunting, given

Re: [web2py] Scheduler - a second try

2013-11-08 Thread Jonathan Lundell
On 8 Nov 2013, at 5:45 AM, Niphlod wrote: > the only thing is that only syslog-ng is multiprocess-safe, so the standard > logging to file and/or console may incur in problems. That applies to more than the scheduler, though, and using syslog isn't so bad. (Maybe we could add a database-based h

[web2py] Scheduler - a second try

2013-11-07 Thread Johann Spies
It was a year ago that I have tried for 4 days to get the scheduler working. In the end I gave up and decided to use the operating systems's cron to handle scheduled jobs. I am trying again and so far I have no success. Running web2py with D0 I get the message DEBUG:web2py.scheduler.artikel#29