[web2py] Re: Need support for web2py application with over 1 million users.

2015-04-24 Thread Niphlod
Database, Yes. Redis, Nope with lock_Keys. I'd go as far as pointing out that the contrib redis backend is the fastest, less cumbersome and less maintenance needy of the whole set of possibilities. -- Resources: - http://web2py.com - http://web2py.com/book (Documentation) -

[web2py] Re: Need support for web2py application with over 1 million users.

2015-04-24 Thread JorgeH
It would be great to have a blog article written about this topic by the ones who know the most. Or to update the web2py manual book with all these pointers. On Friday, April 24, 2015 at 7:49:18 AM UTC-5, Niphlod wrote: Database, Yes. Redis, Nope with lock_Keys. I'd go as far as pointing out

Re: [web2py] Re: Need support for web2py application with over 1 million users.

2015-04-24 Thread Michele Comitini
There is no general advice... Think you were a doctor receiving a call from an unknown person: I feel bad, can you help me? What would you reply based on that information? Take some vitamins ?? 2015-04-23 19:26 GMT+02:00 Paolo Valleri paolo.vall...@gmail.com: Mind that if you store session in

[web2py] DB2 database issue with Clob types

2015-04-24 Thread Boris Aramis Aguilar Rodríguez
class pyodbc.Error' ('ODBC data type -99 is not supported. Cannot read column LDTEXT.', 'HY000') Hi, Im currently accesing a DB2 database (readonly) that is used by another application, my connection string is as follows: db_tivoli =

Re: [web2py] Re: Scheduler exception when setting immediate=True: 'Deadlock found when trying to get lock...'

2015-04-24 Thread Osman Masood
Nope...as proven earlier, this is a web2py timing issue. It's been occurring randomly in production for some time now. I have a try/except surrounding it, but the sucky part is that since the transaction fails, you lose all the commits in the whole request. On Friday, February 27, 2015 at

[web2py] Re: Janrain: NEW OpenID connect

2015-04-24 Thread Dmitry Ermolaev
I use only email.registration it simple - not logins not passwords! example: http://lite.cash/bets/man/greet def set_session(sess, man): sess.man_id = man.id sess.man_name = man.name # confirm - подтверждение по почте получено def greet_conf(): key = request.args(0) if not key:

[web2py] Re: LOAD() problem: Yes, I am bungling the world's smallest web2py app.

2015-04-24 Thread Dmitry Ermolaev
index.html: {{extend 'layout.html'}} {{=LOAD('default', 'showtasks',ajax=True)}} In default.py def index(): return dict() # will take no view def showtasks(): return SQLFORM.grid(db.task,user_signature=False) четверг, 23 апреля 2015 г., 20:00:50 UTC+3 пользователь Tom Campbell

Re: [web2py] LOAD component inside component

2015-04-24 Thread Dmitry Ermolaev
yes вторник, 21 апреля 2015 г., 22:17:38 UTC+3 пользователь Dave S написал: On Tuesday, April 21, 2015 at 6:14:35 AM UTC-7, Dmitry Ermolaev wrote: before to reload a sub-LOAD need stop it auto-loading - for preventing from double loads of sub-LOAD component Does this replace the

Re: [web2py] Need support for web2py application with over 1 million users.

2015-04-24 Thread Vinicius Assef
On 24 Apr 2015, at 09:31, Michele Comitini michele.comit...@gmail.com wrote: There is no general advice… Yes, there are many, as some were already pointed out in this thread, others not yet: - optimize DB queries; - create indexes in your BD; - session.forget() when not necessary; - avoid

[web2py] How to remove application name from urls generated from a scheduled task?

2015-04-24 Thread Lisandro
I'm using parametric based routes for mapping each domain with its corresponding app. Because of this, when I use the URL() helper, the application name isn't included in the resulting url, and that is perfect for me. However, inside a scheduled task, the URL() helper generates urls including

[web2py] ghost scheduler workers can never be deleted even the web2py-schduler is stopped.

2015-04-24 Thread Pengfei Yu
Hi I am using web2py scheduler. I have a project saved in git and then I cloned this git repository to another new instance (server). In the new server, five scheduler workers (known from the ip address in the worker_name) from the old server are always there. When never I delete them, they