openerp-india team mailing list archive

[Florent <1080822@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

I discovered on our production server (running 6.1) that the same job is
run twice, and it gives weird errors.

In our case, the job is retrieving mails on a POP3 server.
While the 1st thread deletes mails, the second one is trying to read them.
At the end it gives random errors.

This bug was already present in version 5.0, and I worked out a patch to deal with it.
Unfortunately the patch was never merged upstream.

Details.
The request "UPDATE ir_cron SET nextcall=%s" is done at the end of the job.
And there's no protection to avoid the job being launched twice.

For example if there's another "cron wake_up" a short time after the
first then the job which is already running is launched again (because
the nextcall value is still in the past).

I will fix it in our codebase and contribute a patch.
I don't know if this is fixed in 7.0.

** Affects: openobject-server
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of OpenERP
Indian Team, which is subscribed to OpenERP Server.
https://bugs.launchpad.net/bugs/1080822

Title:
  the ir.cron scheduler spawns twice the same job

Status in OpenERP Server:
  New

Bug description:
  I discovered on our production server (running 6.1) that the same job
  is run twice, and it gives weird errors.

  In our case, the job is retrieving mails on a POP3 server.
  While the 1st thread deletes mails, the second one is trying to read them.
  At the end it gives random errors.

  This bug was already present in version 5.0, and I worked out a patch to deal with it.
  Unfortunately the patch was never merged upstream.

  Details.
  The request "UPDATE ir_cron SET nextcall=%s" is done at the end of the job.
  And there's no protection to avoid the job being launched twice.

  For example if there's another "cron wake_up" a short time after the
  first then the job which is already running is launched again (because
  the nextcall value is still in the past).

  I will fix it in our codebase and contribute a patch.
  I don't know if this is fixed in 7.0.

To manage notifications about this bug go to:
https://bugs.launchpad.net/openobject-server/+bug/1080822/+subscriptions