| 00:09.17 | *** join/#tomcat jasonb_ (i=noneoyer@adsl-66-124-73-250.dsl.sntc01.pacbell.net) |
| 00:30.28 | *** join/#tomcat Hellaenergy (n=Hellaene@c-75-72-155-242.hsd1.mn.comcast.net) |
| 00:31.40 | *** mode/#tomcat [+o Hellaenergy] by ChanServ |
| 00:32.34 | *** topic/#tomcat by Hellaenergy -> Stable versions: 6.0.10, 5.5.23 and 4.1.36. Newbies use the official binary from tomcat.apache.org, or an RPM package from http://www.webdroid.org:8080/archives/tomcat-package. Check your Tomcat logs before you ask for an answer. SLOW MOTION CHANNEL (we all have jobs & kids): Ask your question, including your TC, Java, & OS versions, then wait; check back often for answers. |
| 01:52.31 | *** mode/#tomcat [-o Hellaenergy] by Hellaenergy |
| 01:57.37 | *** join/#tomcat codeshepherd (n=codeshep@bb121-7-80-240.singnet.com.sg) |
| 03:00.00 | *** join/#tomcat _syphilis_ (n=river@fuchsia.knams.wikimedia.org) |
| 03:00.32 | _syphilis_ | i'm running tomcat embedded into eclipse for development, and occasionally i get "OutOfMemoryError: PermGen space". what kind of resource could i be leaking that would cause this error? |
| 03:09.04 | Hiryu | where can I find info on configuring mod_jk for use with SSL? |
| 03:09.13 | Hiryu | memory resources? |
| 03:10.15 | _syphilis_ | well yes.. what i meant was, is there any particular resource that consumes "PermGen space" as opposed to some other kind of space? |
| 03:21.32 | Hiryu | well, dunno :( |
| 03:34.58 | lyken | mod_jk is painful |
| 03:34.59 | lyken | heh |
| 03:35.29 | lyken | _syphilis_: no idea, id suggest going through your code and identifying bits of code that 'might' consume a lot of memory |
| 03:35.35 | lyken | big db calls and result sets |
| 03:37.27 | Hiryu | got it working |
| 03:37.33 | Hiryu | just wish it was documented better |
| 03:39.44 | Hiryu | oh this is interesting |
| 03:39.57 | Hiryu | it is using apache's ssl to connect to 8080 on the tomcat box |
| 03:40.04 | Hiryu | rather than using the tomcat ssl |
| 03:40.13 | Hiryu | I'm not sure how I feel about that |
| 03:40.13 | Hiryu | hahaha |
| 03:40.23 | Hiryu | maybe that's just how mod_jk is supposed to do ssl? |
| 04:29.00 | *** join/#tomcat jasonb_ (i=noneoyer@adsl-66-124-73-250.dsl.sntc01.pacbell.net) |
| 05:45.49 | *** join/#tomcat caoer (n=mlh@221.12.184.4) |
| 06:07.16 | *** join/#tomcat prgrmr (n=prgrmr@bzq-88-155-122-85.red.bezeqint.net) |
| 07:29.51 | *** part/#tomcat _syphilis_ (n=river@fuchsia.knams.wikimedia.org) |
| 08:47.21 | *** join/#tomcat cforce (i=cforce@dslb-084-056-182-212.pools.arcor-ip.net) |
| 08:47.40 | *** join/#tomcat cforce (i=cforce@dslb-084-056-182-212.pools.arcor-ip.net) |
| 08:48.17 | cforce | hi |
| 09:40.52 | *** join/#tomcat waspdigger (n=waspdigg@202.86.204.197) |
| 09:42.52 | kraut | moin |
| 09:58.35 | *** join/#tomcat Blissex2 (n=pcg@82-69-39-138.dsl.in-addr.zen.co.uk) |
| 10:34.31 | cforce | is there a framework for c# to access servlets and/or ejb's? |
| 10:52.36 | *** join/#tomcat gregor_k (n=a@p54a1ae0f.dip0.t-ipconnect.de) |
| 13:13.43 | *** join/#tomcat fzlogik (n=fuzelogi@dhcp-152-78-61-203.ecs.soton.ac.uk) |
| 13:16.10 | *** part/#tomcat codeshepherd (n=codeshep@bb121-7-80-240.singnet.com.sg) |
| 13:18.43 | *** join/#tomcat codeshepherd (n=codeshep@bb121-7-80-240.singnet.com.sg) |
| 13:19.57 | *** join/#tomcat codeshepherd (n=codeshep@bb121-7-80-240.singnet.com.sg) |
| 13:35.42 | *** join/#tomcat dvayanu (n=another@ppp-82-135-70-116.dynamic.mnet-online.de) |
| 13:45.14 | *** part/#tomcat codeshepherd (n=codeshep@bb121-7-80-240.singnet.com.sg) |
| 13:45.56 | evosh | Good morning. |
| 13:47.14 | evosh | What is an equivalent of org.apache.catalina.connector.http.HttpConnector of Tomcat 4.x in Tomcat 5.5? |
| 14:03.22 | Blissex2 | evosh: it does not a particular name. |
| 14:04.37 | Blissex2 | evosh: http://pastebin.ca/429822 |
| 14:17.31 | evosh | Huh! |
| 14:17.54 | evosh | I am not looking for configurations but rather the actual class for the source code. |
| 14:35.17 | *** join/#tomcat nightfox (n=automato@125.212.63.132) |
| 14:35.46 | *** part/#tomcat nightfox (n=automato@125.212.63.132) |
| 14:50.29 | *** join/#tomcat a4akb (n=Akbara@62.215.156.215) |
| 15:16.53 | Blissex2 | evosh: I think it is still called Coyote |
| 15:26.49 | *** join/#tomcat simie (n=simone@p54819b21.dip0.t-ipconnect.de) |
| 15:26.51 | simie | hi |
| 15:29.02 | simie | can I somehow, from within a running webapp, get Access to the classpath that my webapp uses? Reason: I issue a (shell?) command java via runtime.exec() and need to have the jars in the classpath that are in the WEB-INF7lib folder |
| 15:29.10 | simie | any way to do this? |
| 16:10.50 | *** join/#tomcat yell0w (n=yellow@unaffiliated/yellow) |
| 16:14.19 | *** part/#tomcat a4akb (n=Akbara@62.215.156.215) |
| 16:20.18 | *** join/#tomcat ramza3 (n=ramza3@c-24-99-72-157.hsd1.ga.comcast.net) |
| 16:28.57 | *** join/#tomcat cforce (i=cforce@dslb-084-056-182-212.pools.arcor-ip.net) |
| 16:35.50 | cforce | <PROTECTED> |
| 16:36.06 | cforce | the directory exists, so anyway, what is the problem here? |
| 18:01.17 | Blissex2 | cforce: the webapp is not there? |
| 18:03.30 | cforce | the directory "apache-tomcat-6.0.10\webapps\host-manager" exists |
| 18:06.18 | *** join/#tomcat slaq (n=y@port-212-202-38-87.dynamic.qsc.de) |
| 18:06.30 | slaq | hi ppl |
| 18:07.13 | slaq | is it possible to call to a jsp from inside my java code (not a servlet) and receive the response? |
| 18:07.30 | slaq | i'd like to use jsps as flexible code modules for input validation |
| 18:07.51 | slaq | but i have no idea how to access them outside the context of a servlet |
| 19:22.08 | dvayanu | slaq: that seems silly :-) |
| 19:22.36 | dvayanu | slaq: but i assume you can forward to a jsp and let the jsp forward to something else |
| 19:22.40 | dvayanu | from a servlet |
| 19:23.15 | dvayanu | errm, not a servlet, still you can call jsp's service method with fak response/request objects |
| 19:53.41 | jasonb | Yeah, the answer is yes, it ends up being a Java class so it is possible to do. |
| 20:06.36 | *** join/#tomcat jasonb_ (i=noneoyer@adsl-66-124-73-250.dsl.sntc01.pacbell.net) |
| 20:37.31 | *** join/#tomcat seevash (n=jixjax@user-1121ija.dsl.mindspring.com) |
| 20:37.31 | *** join/#tomcat a4akb (n=Akbara@62.215.156.215) |
| 20:44.20 | seevash | Hey, welcome back seevash. |
| 21:54.08 | *** part/#tomcat a4akb (n=Akbara@62.215.156.215) |
| 22:00.25 | slaq | dvayanu, can you elaborate on what you said earlier? |
| 22:00.36 | slaq | how can i access a jsp, like how do i even figure out an instance? |
| 22:00.54 | slaq | i have drilled down the jsp processing down to DispatchRequest.java |
| 22:01.05 | slaq | but this stuff doesn't seem like it's designed to be hooked into ;) |
| 22:02.06 | slaq | calling the service method would be good enough for me i guess |
| 22:02.18 | slaq | i'm looking for a simple way to plug in dynamic pieces of code into my app |
| 22:03.22 | slaq | have been using janino for a while but it feels a bit clumsy to me, jsp seems like a more natural choice (simply because its already there) |
| 22:17.21 | dvayanu | hey slaq |
| 22:17.45 | dvayanu | slaq: i think you are trying the wrong tool |
| 22:21.02 | dvayanu | 35 |
| 22:24.56 | slaq | dvayanu, hrm |
| 22:25.09 | slaq | well, do you have a better idea? :) |
| 22:28.03 | slaq | i think it would be quite neat in my case but only if there is a way to wrap it up in a simple "public String doJsp( String filename )" ... |
| 22:42.26 | *** join/#tomcat sunbug (n=sunbug@absolutlinux.no) |
| 22:49.19 | *** join/#tomcat thatguy (n=thatguy@207-38-162-187.c3-0.wsd-ubr5.qens-wsd.ny.cable.rcn.com) |
| 23:01.50 | dvayanu | slaq: what exactly are your requirements |
| 23:03.57 | *** join/#tomcat evosh (n=jixjax@user-1121ija.dsl.mindspring.com) |
| 23:23.10 | *** join/#tomcat seevash (n=jixjax@user-1121gp4.dsl.mindspring.com) |
| 23:29.25 | jasonb | slaq: You have to remember that JSP is designed to run inside a servlet container that has a JSP engine. You may be able to invoke the service() method of the JSP servlet, but the JSP servlet will likely expect the full web container environment. Would you be calling service() from within a webapp? Or, from a Java program that isn't running any servlet container? |
| 23:40.32 | slaq | jasonb, from within a webapp |
| 23:41.26 | slaq | jasonb, i'm prepared to fake ServletRequest and ServletResponse if i need to, but if it needs much more then i guess i'd rather look for another solution |
| 23:41.35 | jasonb | slaq: Some reason why you cannot use the request dispatcher? |
| 23:42.12 | slaq | jasonb, well, how would i invoke the dispatcher from, say, a servlet and collect the response (the output JspWriter) |
| 23:43.19 | jasonb | slaq: Another way you could do this is to include an HTTP client in your webapp, like Commons-HttpClient, and use that to make another request to the servlet container to invoke the JSP of your choice. |
| 23:43.37 | jasonb | slaq: It isn't real efficient, but it does the job. |
| 23:43.41 | slaq | jasonb, well, i'd like to keep the overhead low. |
| 23:44.02 | jasonb | slaq: Okay, so would you then be open to making your webapp have hard dependencies on Tomcat internals? |
| 23:44.21 | slaq | jasonb, that would be resin internals but i guess yes ;) |
| 23:44.35 | jasonb | slaq: Well, I don't know how to do what I'm thinking for Resin. |
| 23:44.44 | slaq | hmmm |
| 23:44.53 | jasonb | slaq: You're supposed to use Tomcat. :) |
| 23:45.01 | slaq | but resin is faster :) |
| 23:45.04 | slaq | anyways, another thought |
| 23:45.06 | jasonb | slaq: No it isn't. |
| 23:45.48 | slaq | last time i benched it was |
| 23:46.16 | jasonb | slaq: What I was going to suggest is: you could make your webapp a priviliged webapp so that it has access to Tomcat internal objects, then make the webapp schedule its own request on Tomcat to the JSP, and receive back the response. |
| 23:46.33 | jasonb | slaq: Which version of Tomcat was it that you benchmarked, and what was your connector configuration? |
| 23:46.49 | slaq | jasonb, 5.something and there was no connector |
| 23:47.05 | jasonb | Then you're not benchmarking Tomcat. |
| 23:47.16 | jasonb | This runs over the web, riiight? |
| 23:47.24 | slaq | oh sorry |
| 23:47.28 | slaq | the plain http connector |
| 23:47.44 | slaq | i meant there was no addtl webserver |
| 23:47.55 | jasonb | Well, 5.something probably means 5.0, which wasn't a good performing version of Tomcat. |
| 23:48.23 | jasonb | Tomcat 6.0 is very fast. |
| 23:48.33 | slaq | hm, i dont want to get into container wars here :) |
| 23:48.37 | jasonb | Faster than Apache httpd. |
| 23:48.48 | slaq | erm, apache is slow but not *that* slow ;) |
| 23:48.56 | slaq | we |
| 23:49.00 | slaq | we're talking java after all.. |
| 23:49.12 | jasonb | Java can be faster than C. |
| 23:49.20 | jasonb | In some cases, it is. |
| 23:49.29 | slaq | i don't want to get into language wars either ;) |
| 23:49.34 | jasonb | And, I have all the benchmark data to prove my point. |
| 23:49.36 | slaq | i just wonder why the jsp processing is so closed |
| 23:49.49 | jasonb | "closed"? |
| 23:50.09 | slaq | yea, we have dynamic compiling and classloading at our fingertips but are not supposed to use it |
| 23:50.18 | jasonb | Wars aside, I just wanted to know what was fastest, and why, so I gathered the information, and now I know the answer. |
| 23:50.47 | jasonb | You can use dynamic compiling and classloading for this all you want to. |
| 23:51.25 | jasonb | The dynamic compilation isn't really designed to be triggered by servlet code, which is why there is no API for doing that. |
| 23:51.36 | jasonb | (in the Servlet API, that is) |
| 23:51.39 | slaq | in my case it would be useful to do just that :) |
| 23:51.56 | jasonb | Well, you can, if you could use Tomcat internal objects. |
| 23:52.37 | slaq | well, if i have to use internal objects i guess i'll have to pest the resin ml |
| 23:52.51 | jasonb | They will tell you the same thing. |
| 23:52.54 | slaq | or rather solve my problem in a more cumbersome way that doesnt involve jsps |
| 23:53.02 | slaq | yes but they'd be able to tell me which objects to use in resin ;) |
| 23:53.23 | jasonb | You're better off if you do not have to invoke the JSP. Any code you want to invoke should really be in a tag library or something similar anyway. |
| 23:53.54 | slaq | well, it's about duplication. the validation logic that i'd like to put there would be useful both when triggered internally and when called externally (via http) |
| 23:54.13 | slaq | i don't understand why the jsp processing isn't just implemented as a regular pipeline that i can hook into at various stages |
| 23:55.31 | slaq | i mean, it isn't easy to get the recompilation and class-reloading right in an own impl, so why is the one known working impl closed for reuse :\ |
| 23:55.47 | slaq | not a beef with tomcat btw, more with the guys who write the specs |
| 23:57.05 | jasonb | Well, the request handling in the servlet spec is indeed a sort of pipeline, but it is not so general purpose that it handles all use cases like yours in addition to the ones they were trying to support. |
| 23:57.25 | jasonb | Adding more features would complicate all servlet container implementations, and would probably also cost some performance as well. |
| 23:57.31 | slaq | hum |
| 23:58.22 | slaq | i can't imagine a simple public String callJsp( HashMap parameterMap, String queryString, etc .. ); would impose any overhead or more complication |
| 23:58.36 | slaq | erm, make the return type OutputStream if you so prefer |
| 23:59.19 | slaq | put that into tomcat and i'll actually look into switching back ;) |
| 23:59.31 | slaq | maybe v6 is worth benchmarking again, resin ain't free |