Comment 22 for bug 717345
Revision history for this message
| Martin Pool (mbp) wrote Re: [Bug 717345] Re: Updates to SFTP server leak file handles in use_forking_server mode | #22 |
© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
73416ba
(Get the code!)
On 17 February 2011 09:27, Robert Collins <email address hidden> wrote:
> "
> Note that Andrew has already started work on allowing us to run multiple
> Conch processes (and associated forking processes), so that we can
> handle no-downtime deployments anyway. Which also helps us with high
> availability, etc."
>
> what does this mean? I wouldn't want to run > 1 conch per forking
> service, or vice versa. If hes simply talking about having multiple
> conch processes with an haproxy front end, we're already ready to do
> that; the thing that andrew was working on for a status page is about
> graceful handoffs when doing nodowntime deploys (and thats pretty
> important in the long-running connection model of bzr).
I think it's just that John sees this feature as "being worked on"
whereas Robert sees it as "ready to deploy it and to do follow on
work." (bug 702024, or RT 40480). It would let us if we wanted
limit each conch to just 256 connections. I don't know of any plans
to have other than one conch per forking server.