Bug#425844: cyrus reload problem More info

2007-06-16 Thread Ross Boylan
On Sat, 2007-06-16 at 15:51 +0100, Benjamin Seidenberg wrote:
 retitle 425844 cyrus-imapd-2.2: cyrmaster dies on reload/force-reload
 thanks
 
 Reload is the same action.
AFAIK cyrmaster is not directly responsible for listening on the lmtp
socket.  Why is that affected?  Does the reload shutdown the imapd
process as well?

Your subject was More info.  Is that a request for more info?  There
was no more info tag added, so I'm guessing it wasn't.

P.S. Bug reports to cyrus are still generating bounce-back like
Your mail to 'Pkg-Cyrus-imapd-Debian-devel' with the subject

Bug#429164: cyrus-imapd-2.2: reload seems to block unix domain
socket deliveries

Is being held until the list moderator can review it for approval.

The reason it is being held:

Post by non-member to a members-only list

My original report was from a different address than this message; I
think this message is subscirbed to the list.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#425844: cyrus reload problem More info

2007-06-16 Thread Benjamin Seidenberg
Ross Boylan wrote:
 On Sat, 2007-06-16 at 15:51 +0100, Benjamin Seidenberg wrote:
 retitle 425844 cyrus-imapd-2.2: cyrmaster dies on reload/force-reload
 thanks

 Reload is the same action.
 AFAIK cyrmaster is not directly responsible for listening on the lmtp
 socket.  Why is that affected?  Does the reload shutdown the imapd
 process as well?

It calls a function reread_conf. I'm looking at it right now.

 
 Your subject was More info.  Is that a request for more info?  There
 was no more info tag added, so I'm guessing it wasn't.

No, it meant I was adding (very little) info.

 
 P.S. Bug reports to cyrus are still generating bounce-back like
 Your mail to 'Pkg-Cyrus-imapd-Debian-devel' with the subject
 
 Bug#429164: cyrus-imapd-2.2: reload seems to block unix domain
 socket deliveries
 
 Is being held until the list moderator can review it for approval.
 
 The reason it is being held:
 
 Post by non-member to a members-only list
 
 My original report was from a different address than this message; I
 think this message is subscirbed to the list.


It was a different email than you normally used, I had already approved
it and added it to the auto-approve list.

Benjamin




Bug#425844: cyrus reload problem More info

2007-06-16 Thread Ross Boylan
On Sat, 2007-06-16 at 17:20 +0100, Benjamin Seidenberg wrote:
 Ross Boylan wrote:
  On Sat, 2007-06-16 at 15:51 +0100, Benjamin Seidenberg wrote:
  retitle 425844 cyrus-imapd-2.2: cyrmaster dies on reload/force-reload
  thanks
 
  Reload is the same action.
  AFAIK cyrmaster is not directly responsible for listening on the lmtp
  socket.  Why is that affected?  Does the reload shutdown the imapd
  process as well?
 
 It calls a function reread_conf. I'm looking at it right now.
My logs show cyrmaster received a SIGHUP when I did the reload, if that
helps.
 
  
  Your subject was More info.  Is that a request for more info?  There
  was no more info tag added, so I'm guessing it wasn't.
 
 No, it meant I was adding (very little) info.
 
It was very helpful to me; I wasn't sure if reload and force-reload were
the same.  Probably it's obvious if one reads the fine code.
  
  P.S. Bug reports to cyrus are still generating bounce-back like
  Your mail to 'Pkg-Cyrus-imapd-Debian-devel' with the subject
  
  Bug#429164: cyrus-imapd-2.2: reload seems to block unix domain
  socket deliveries
  
  Is being held until the list moderator can review it for approval.
  
  The reason it is being held:
  
  Post by non-member to a members-only list
  
  My original report was from a different address than this message; I
  think this message is subscirbed to the list.
 
 
 It was a different email than you normally used, I had already approved
 it and added it to the auto-approve list.
 
I didn't mean the comment so much for me as for others who submit bugs
against cyrus.  They may find it confusing or off-putting to receive
those bounce messages.  If there's a way to prevent them, that would be
good.  I'm pretty sure other projects also forward their bugs to a
maintainer list; I'm not sure why cyrus seems to behave differently.
Ross



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]