RE: mod_jk startup

2002-06-29 Thread Mladen Turk
> -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] > Sent: 29. lipanj 2002 20:08 > To: Tomcat Developers List > Subject: RE: mod_jk startup > > > On Sat, 29 Jun 2002, Mladen Turk wrote: > > > > Also, why a new jkS

RE: mod_jk startup

2002-06-29 Thread costinm
On Sat, 29 Jun 2002, Mladen Turk wrote: > > Also, why a new jkStatus method instead of setAttribute() ? > > > > Need the static call, cause its called from TomcatStarter, and since I > don't need to pass nothing fancy, that was the simplest thing I can > think of. Well, it's ok - but long term

RE: mod_jk startup

2002-06-29 Thread Mladen Turk
> -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] > Sent: 29. lipanj 2002 16:59 > To: Tomcat Developers List > Subject: Re: mod_jk startup > > > > > > That means that all the calls will be locked until the > worker fini

Re: mod_jk startup

2002-06-29 Thread costinm
On Sat, 29 Jun 2002, Mladen Turk wrote: > All the worker initialization checking is done in the load balancer. +1 ! ( can I say "I was thinking the same" ? ) > The load balancer has now the additional config parameter 'initTimeout' > (default is 1 minute). > > That means that all the calls wi

mod_jk startup

2002-06-29 Thread Mladen Turk
Hi, Some more explanations about the recent patch. All the worker initialization checking is done in the load balancer. The only drawback is that all the calls are disabled until the worker finishes with the initialization (perhaps it could be done inside the worker itself but IMO it will be th