zkn ha scritto:
On 30.07.2008, at 12:40, Stefano Bagnara wrote:
zkn ha scritto:
On 25.07.2008, at 10:50, Stefano Bagnara wrote:
zkn ha scritto:
On 25.07.2008, at 10:12, Stefano Bagnara wrote:
With many mails the db based spool repository is much better.
Change of processor is done by updating few fields in a db record.
Stefano
Thanks Stefano.
Now when we have so many files in the file system is it still
possible to switch to db spool and would that help processing these
emails faster?
I say it would be much more fast, but it highly depends on your
system setup.
In order to move mails the only solution with james 2.3.1 is to move
it to db and then use the FromRepository mailet to retrieve mails
from the file repository to the spool repository.
I'm not sure but it seems to me that if I use FromRepository I will
loose new emails. The mailet is triggered by a new email and then it
starts retrieving emails from the repository to the spool. And the
original email is GHOSTed.
Is that correct?
IIRC there is a <delete> parameter for the FromRepository mailet
I'm looking at the code and <delete> is used only for emails in the
source repository while the new email that triggered the mailet is
unconditionally GHOSTed:
public void service(Mail trigger) throws MessagingException {
trigger.setState(Mail.GHOST);
I might be wrong but if that's the case it doesn't look good to me.
The idea is that you configure a special address for triggering the
fromrepository, and then use a matcher to catch that specific mail.
The mail purpose is only to trigger, so when the triggering start it can
be ghosted.
E.g:
<mailet match="[EMAIL PROTECTED]" class="FromRepository">
Stefano
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
