Hi list, I have tried to reach Bradley (author of mongrel_cluster) by mail, but have not gotten a response. So I''ll try trough this channel: I have ''developed''[1] a small extension to mongrel_cluster that allows selective restart of any one listener in a configuration that contains more than one listener by using a command like mongrel_rails cluster::selectivestart -c config.file -p 3000 This will try to find a listener on port 3000 in the configuration given and start it (selectively). This is super useful in a monit controlled environment, where a rule could now be: start program = "mongrel_rails cluster::selectivestart -C mongrel_cluster.yml -p 3000" Of course, I have a stop action, too. Now the question: Bradley, if this reaches you, are you interested in including this into mongrel_cluster? Or should I just go ahead and release this independently? Is anyone interested in this? best greetings kaspar [1] By using Apple C, Apple V mostly...
On Mon, 18 Dec 2006 09:52:45 +0100 Kaspar Schiess <eule at space.ch> wrote:> Hi list, > > I have tried to reach Bradley (author of mongrel_cluster) by mail, but > have not gotten a response. So I''ll try trough this channel: >Send it to me offlist and I''ll take a look at it. Thanks. -- Zed A. Shaw, MUDCRAP-CE Master Black Belt Sifu http://www.zedshaw.com/ http://www.awprofessional.com/title/0321483502 -- The Mongrel Book http://mongrel.rubyforge.org/ http://www.lingr.com/room/3yXhqKbfPy8 -- Come get help.
hi all, wouldn''t this be really useful for updating production servers as well? instead of taking down all mongrels at once, you could restart one after another. so you have zero downtime, and depeding on how smart your script is even cancel restart of mongrel n+1 if mongrel n hasn''t come up successfully. cheers, phillip --- Am 18.12.2006 um 10:53 schrieb Zed A. Shaw:> On Mon, 18 Dec 2006 09:52:45 +0100 > Kaspar Schiess <eule at space.ch> wrote: > >> Hi list, >> >> I have tried to reach Bradley (author of mongrel_cluster) by mail, >> but >> have not gotten a response. So I''ll try trough this channel: >> > > Send it to me offlist and I''ll take a look at it. Thanks. > > -- > Zed A. Shaw, MUDCRAP-CE Master Black Belt Sifu > http://www.zedshaw.com/ > http://www.awprofessional.com/title/0321483502 -- The Mongrel Book > http://mongrel.rubyforge.org/ > http://www.lingr.com/room/3yXhqKbfPy8 -- Come get help. > _______________________________________________ > Mongrel-users mailing list > Mongrel-users at rubyforge.org > http://rubyforge.org/mailman/listinfo/mongrel-users > > >------------- Phillip Oertel Manager Engineering phillip.oertel at xing.com Open Business Club AG, XING Gaensemarkt 43, 20354 Hamburg, Germany Tel +49 40 41 91 31-754, Fax +49 40 41 91 31-11 Please join my network on XING: http://www.xing.com/profile/Phillip_Oertel This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorised copying, disclosure or distribution of the material in this e-mail is strictly forbidden and may be unlawful.
On Mon, 18 Dec 2006 11:43:52 +0100 Phillip Oertel <phillip.oertel at openbc.com> wrote:> hi all, > > wouldn''t this be really useful for updating production servers as well? > instead of taking down all mongrels at once, you could restart one > after another. > so you have zero downtime, and depeding on how smart your script is > even cancel restart of > mongrel n+1 if mongrel n hasn''t come up successfully.Problem is this gets ultra complex and everyone wants it done differently. Most folks who want this simply write their own version and get exactly what they want. A general solution seems to be a whopper of a task (trust me I''ve tried a few times). If you can write something up and post it out then go for it. Mongrel gives folks the tools to do custom management processes with this in mind. -- Zed A. Shaw, MUDCRAP-CE Master Black Belt Sifu http://www.zedshaw.com/ http://www.awprofessional.com/title/0321483502 -- The Mongrel Book http://mongrel.rubyforge.org/ http://www.lingr.com/room/3yXhqKbfPy8 -- Come get help.
Am 18.12.2006 um 12:14 schrieb Zed A. Shaw:> On Mon, 18 Dec 2006 11:43:52 +0100 > Phillip Oertel <phillip.oertel at openbc.com> wrote: > >> hi all, >> >> wouldn''t this be really useful for updating production servers as >> well? >> instead of taking down all mongrels at once, you could restart one >> after another. >> so you have zero downtime, and depeding on how smart your script is >> even cancel restart of >> mongrel n+1 if mongrel n hasn''t come up successfully. > > Problem is this gets ultra complex and everyone wants it done > differently. Most folks who want this simply write their own > version and get exactly what they want. A general solution seems > to be a whopper of a task (trust me I''ve tried a few times). > > If you can write something up and post it out then go for it. > Mongrel gives folks the tools to do custom management processes > with this in mind.very cool :-) so i''ll try to write up something for our environment, and depending on how much i like it i''ll post it. we don''t need that urgently yet, so it might be some time. cheers, phillip
> I have ''developed''[1] a small extension to mongrel_cluster that allows > selective restart of any one listener in a configuration that contains > more than one listener by using a command like > > mongrel_rails cluster::selectivestart -c config.file -p 3000 > > This will try to find a listener on port 3000 in the configuration given > and start it (selectively).> > Now the question: Bradley, if this reaches you, are you interested in > including this into mongrel_cluster? Or should I just go ahead and > release this independently? Is anyone interested in this?Hi Kaspar: I just found your mail buried deep in my fluxura box. Sorry about that! I guess my main question is why would you use mongrel_cluster instead of the Mongrel configuration file option (-C) and specify the port on the command line like ''mongrel_rails start -C /path/to/your/file -p 3000''? Regards, Bradley Taylor Rails Machine Virtual Clusters for Web Deployments http://railsmachine.com
Hi Brad> I just found your mail buried deep in my fluxura box. Sorry about that! > > I guess my main question is why would you use mongrel_cluster instead of > the Mongrel configuration file option (-C) and specify the port on > the command line like ''mongrel_rails start -C /path/to/your/file -p 3000''?The way I understand this command is that it will only ever allow me to start one instance at once. (side note: I am not sure as to if -C does override any other command line options or not.) What I want to be able to do is manage a couple of mongrels all at once using mongrel cluster::start/stop/restart _and_ selectively manipulate mongrels in that cluster, while not breaking pid-file handling, etc.. of mongrel_cluster. If you tell me that I can get all of this with the existing code, I will wield the quick ''rm'' on all the new code. But I cannot see how you would accomplish clustering and individual restarting just trough mongrel start/stop/restart. Thanks for getting back to me! kaspar