Lucas Zinato Carraro wrote:
WC There is, however, no way to keep the connection to backends
WC open beyond between each user session. If you have a lot
WC of short lived IMAP connections, you're probably interested
WC in up-imapproxy.
LZC I believe that most customers will use Thunderbird and
On 22 Jun 2010, at 00:26, Lucas Zinato Carraro wrote:
proxyd_disable_mailbox_referrals will be replaced to other
parameter in new branch ( 2.4 )?
Either that or proxyd_disable_mailbox_referrals will be ported forward.
IDLE imap command is compatible in Cyrus Murder enviromment ??
It
Thanks for all answers,
As Wes points out, proxyd process churn is relatively expensive.
Any metric to sizing a solution ?
ex: proxyd memory and cpu by connection
I expect 20.000 simultaneous connections in frontends ( 3 or 5 machines )
10.000 using IMAPs and 10.000 using IMAP.
An
On Tue, 22 Jun 2010, Lucas Zinato Carraro wrote:
Thanks for all answers,
As Wes points out, proxyd process churn is relatively expensive.
Any metric to sizing a solution ?
ex: proxyd memory and cpu by connection
I expect 20.000 simultaneous connections in frontends ( 3 or 5
Another Question:
1 proxyd = 1 client connection ? Or 1 proxyd handles several connections ?
This parameter can be configured ?
Any metric to sizing IO Use for frontends and backends ??
Regards
On Tue, Jun 22, 2010 at 2:48 PM, Andrew Morgan mor...@orst.edu wrote:
On Tue, 22 Jun 2010,
I have some questions about a possible Cyrus Agregation ( Murder )
implementation in my enviromment.
My envirommet;
--
- Internet Firewall
-- 3 Cyrus FrontEnd ( DMZ )
- Internal DMZ Firewall
-- 3 Backends + 1 Mupdate.
On 21 Jun 2010, at 20:56, Lucas Zinato Carraro wrote:
I find a parameter to imapd.conf: proxyd_disable_mailbox_referrals.
Only this parameter is necessary in frontend to complete disable
referall ?
More or less (that option doesn't appear to be in the 2.4
branch) There's also
-
I find a parameter to imapd.conf: proxyd_disable_mailbox_referrals.
Only this parameter is necessary in frontend to complete disable referall
?
More or