"Johnny Kewl" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] > > --------------------------------------------------------------------------- > HARBOR: http://coolharbor.100free.com/index.htm > The most powerful application server on earth. > The only real POJO Application Server. > Making the Java dream come true. > --------------------------------------------------------------------------- > > > Thanks Filip, I see this is your baby, clearly a guru.. > I have missed the history here and actually didnt even realize there were > competing specs. > I now see what the technology is about, the joggling of blocking and > events, threads are cool, but blocking is easy and they > tossing the 2 concepts into various combinations... > The practical implementation I still struggle with, maybe one day a need > will fall onto my desk, and Comet will jump up as a solution.
Well, Comet is good when you have the client frontend tightly bound to the server backend. For example, the (simplistic) Chat example Servlet that ships with TC6, or, say an Applet that shows real-time stock prices. In these cases, changes happen on the server-side that the client isn't aware of. But that doesn't matter, since the server can push the updates to the client without the client having to constantly poll for new data. Comet lets your app scale much much higher in these types of cases, since you aren't wasting threads and bandwidth with the client asking "anything new?" and the servlet responding "not yet". > I also realize that my post to Bill in terms of this technology is way > off, but to me extreme control like this, if made slightly more generic, > ie not just Http... would mean Tomcat is also applicable to extreme > solutions... > I can hear Christopher, saying "please no, questions like we have a video > server in one Servlet, and Ldap server in another... how can we"... ha ha > Comet works (as implemented in Tomcat) by wrapping the actual protocol in HTTP as a wire protocol. However, the Coyote abstraction makes TC pretty much agnostic to wire protocols as well :). > ..... thanks very much.... Tomcat is simply the best... > > ----- Original Message ----- > From: "Filip Hanik - Dev Lists" <[EMAIL PROTECTED]> > To: "Tomcat Users List" <users@tomcat.apache.org> > Sent: Saturday, February 23, 2008 10:47 PM > Subject: Re: NIO Java or TC - What are u using it 4? > > >> Johnny Kewl wrote: >>> >>> ----- Original Message ----- From: "Filip Hanik - Dev Lists" >>> <[EMAIL PROTECTED]> >>> To: "Tomcat Users List" <users@tomcat.apache.org> >>> Sent: Friday, February 22, 2008 5:59 PM >>> Subject: Re: NIO Java or TC - What are u using it 4? >>> >>> >>>> you're subject has an "or" in it, what are you actually asking? >>> >>> What kind of projects make one want to move away from normal >>> HttpServlets and start >>> using the NIO stuff that TC has... I understand that it gives one all >>> sorts of async control >>> over the servlet... but I just cant think of examples of when one would >>> use it. >>> Occasionally I see a question or two here, and I'm wondering what >>> practical things >>> designers are using these "power" sockets for. >>> >>> There are lots of articles like this one.... >>> http://www.ibm.com/developerworks/library/j-nioserver/ >> ok, well most likely the servlet spec will one of these days introduce >> non blocking IO as well, but it will be while, it's a very complex >> programming model, and in most cases one doesn't need it. if you want >> async servlets (ie, decouple threads from the connection) then as Bill >> mentioned, Comet will do an excellent job. >> switching to something like the link above, you're now dealing with >> software outside the spec, which in some cases is not a problem, but in >> other cases can have consequences for a growing project, or a project >> that needs to be integrated elsewhere. >> >> >>> >>> They espouse the "power" side... but no concrete examples of >>> applications... >>> Its not like people are all diving at it... so I'm wondering if there >>> are times when its the only way to go... >>> Also in TC there is another APR handler, how does that relate to NIO, >>> when will one want to use any of these >>> over the normal, and when APR and when NIO? >>> Just trying to understand it... practical examples sounds like a good >>> way to start... >>> Thanks... >>> >> >> To give you an example of the different connectors and how they differ, >> in a one-to-one scenario, NIO will always be slower, since it has to do >> "simulated blocking" to accomodate the current servlet spec. When it >> comes to SSL, no java connector compares to the APR/OpenSSL, JSSE is >> simply not optimized enough (yet). >> >> Java Blocking Connector Java Nio Blocking Connector >> APR Connector >> Classname Http11Protocol Http11NioProtocol >> Http11AprProtocol >> Tomcat Version 3.x 4.x 5.x 6.x 6.x 5.5.x 6.x >> Support Polling NO YES YES >> Polling Size N/A Unlimited - Restricted by >> mem Unlimited >> Read HTTP Request Blocking Non Blocking >> Blocking >> Read HTTP Body Blocking Sim Blocking >> Blocking >> Write HTTP Response Blocking Sim Blocking >> Blocking >> SSL Support Java SSL Java SSL OpenSSL >> SSL Handshake Blocking Non blocking >> Blocking >> Max Connections maxThreads See polling size >> See pollsize >> >> >> Filip >> >>>> >>>> Filip >>>> >>>> Johnny Kewl wrote: >>>>> Cant contain my curiousity anymore? >>>>> What practical applications is this been used for? >>>>> I see theres a lot of control over the socket etc... but what are you >>>>> guys actually doing with it? >>>>> Trying to get rule of thumb for when one would move to it? >>>>> >>>>> >>>>> --------------------------------------------------------------------------- >>>>> >>>>> HARBOR: http://coolharbor.100free.com/index.htm >>>>> The most powerful application server on earth. >>>>> The only real POJO Application Server. >>>>> Making the Java dream come true. >>>>> --------------------------------------------------------------------------- >>>>> >>>>> ------------------------------------------------------------------------ >>>>> >>>>> >>>>> No virus found in this incoming message. >>>>> Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: >>>>> 269.20.9/1293 - Release Date: 2/22/2008 9:21 AM >>>>> >>>> >>>> >>>> --------------------------------------------------------------------- >>>> To start a new topic, e-mail: users@tomcat.apache.org >>>> To unsubscribe, e-mail: [EMAIL PROTECTED] >>>> For additional commands, e-mail: [EMAIL PROTECTED] >>>> >>>> >>> >>> >>> --------------------------------------------------------------------- >>> To start a new topic, e-mail: users@tomcat.apache.org >>> To unsubscribe, e-mail: [EMAIL PROTECTED] >>> For additional commands, e-mail: [EMAIL PROTECTED] >>> >>> >>> >> >> >> --------------------------------------------------------------------- >> To start a new topic, e-mail: users@tomcat.apache.org >> To unsubscribe, e-mail: [EMAIL PROTECTED] >> For additional commands, e-mail: [EMAIL PROTECTED] >> >> > > > --------------------------------------------------------------------- > To start a new topic, e-mail: users@tomcat.apache.org > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To start a new topic, e-mail: users@tomcat.apache.org To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]