| 00:50.23 | *** join/#tomcat blumber (n=klumber@adsl-69-232-203-172.dsl.pltn13.pacbell.net) |
| 00:59.39 | swente | jasonb: i think i got it. |
| 02:02.37 | *** join/#tomcat vinse_ (n=vinse@ppp-71-139-33-236.dsl.snfc21.pacbell.net) |
| 02:06.40 | jasonb | swente: Quite cool. |
| 02:06.55 | jasonb | swente: Pretty much like we were talking about before? |
| 03:18.02 | swente | jasonb: well, i did not "extend" AccessLogValve -- because it would make it depend on log4j. instead, i made a copy, ripped off all file/rotation-related stuff and built-in, in the first step, just a log4j-logging (to be configured via log4j.properties. a second class derives from that, and included basic programmatic/property-configuration |
| 03:19.10 | swente | jasonb: but, as it contains all pattern-related stuff, it logs quite nicely :) |
| 03:21.36 | *** join/#tomcat lemniskate (n=lemniska@0127bhost185.starwoodbroadband.com) |
| 03:58.52 | jasonb | swente: I'm glad you got it working. |
| 03:59.17 | swente | jasonb: mmh .. i could send you a jar containing all [src+class+doc], it's really just a simple hack |
| 03:59.37 | jasonb | swente: So which classes are there in your implementation, and what are the class names? |
| 04:00.10 | jasonb | From your description, I'd bet that they will not commit it as you designed it. |
| 04:00.34 | swente | jasonb: the package-names (actually) derive from my own domain, i might change if needed. there's a AccessLog4JValve and AccessSyslog4JValve |
| 04:00.49 | jasonb | The reason being that you've duplicated most of the pattern related code.. if I understood your description right. |
| 04:01.01 | swente | anyway, if they dislike it because of it's "style" or something .. well, it's not "needed" to get into tomcat |
| 04:01.07 | swente | yes |
| 04:01.23 | swente | but i had to duplicate most of that code -- or i would have to change the original classes |
| 04:01.50 | jasonb | Well, you can fix your own problem this way, but there are zillions of other people who also want this functionality.. and if you don't get yours into Tomcat, then you're really only fixing yours. |
| 04:02.28 | jasonb | It's okay to change the original classes. Or, subclass and override some things. If by overriding you have to copy huge amounts of code, then that likely means the original should be modified anyway. |
| 04:02.45 | swente | hmm ok. so you think i should rip AccessLogValve to make it pluggable [File|Log4J] ? |
| 04:02.55 | swente | ok |
| 04:03.18 | jasonb | Actually, yes. I wish all of Tomcat's logger classes were more configurable. And, I'm not the only one. |
| 04:04.10 | jasonb | Early on (back in like 2000) I considered revamping AccessLogValve so that it was just the pattern related code, then do a subclass like FileAccessLogValve. I didn't do it only because I was worried about performance. |
| 04:04.25 | jasonb | AccessLogValve is actually a performance-critical section of code. |
| 04:04.39 | jasonb | I worked on it way back then and improved performance... that's why my name is in there. |
| 04:04.53 | swente | :) |
| 04:05.06 | jasonb | I do still think that it should be refactored so that it isn't just always a file logger. |
| 04:05.17 | swente | is subclassing still such a performance penalty nowadays? |
| 04:05.27 | jasonb | It should either be configurable (as one class) to log in different ways, or it should be a few classes. |
| 04:05.39 | swente | hmm . there's also those fastlogvalve-ones |
| 04:06.06 | swente | the performance-hungry might go with "unflexible" ones, but one also might create a more comfortable competitor |
| 04:06.21 | jasonb | There is indeed a performance penalty to subclassing. But, we don't know how big that is until we benchmark it. |
| 04:06.21 | swente | i mean .. not "or", but "as well" :) |
| 04:06.52 | jasonb | The fast*logvalve classes can be ignored... they never went anywhere. |
| 04:07.05 | swente | oh ic:) |
| 04:07.13 | jasonb | Remy really didn't like those implementations. :) |
| 04:08.15 | swente | is there some code-styleguide? |
| 04:09.21 | jasonb | I think maybe one way this could be re-implemented is: isolate all of the file based logging code into its own methods in AccessLogValve, then subclass that into AccessLog4JValve, and in there override all the file based logging methods with log4j ones. Then, subclass AccessLog4JValve with AccessSyslog4JValve, which overrides more stuff to do syslogging. OR make syslog a configurable thing in AccessLog4JValve. |
| 04:09.42 | jasonb | There isn't really a style guide or coding conventions guide.. another reason why lots of patches get rejected. |
| 04:09.53 | swente | hmm ok |
| 04:09.55 | jasonb | I know what the style is.. so if you're in doubt, ask me, or ask on the list. |
| 04:10.03 | jasonb | (the dev list that is) |
| 04:10.16 | swente | ok, you suggestion: basically i would just need to change properties visibility |
| 04:10.27 | jasonb | Yeah. |
| 04:10.52 | swente | is String.format() new from jdk 1.4 or 5.0 |
| 04:10.54 | swente | ? |
| 04:11.01 | jasonb | Also, the method overriding would need to be done in a way that minimizes the performance penalty of subclassing and overriding them. |
| 04:11.17 | jasonb | If it's in the Tomcat source, it's 1.4 and below. |
| 04:11.27 | jasonb | Tomcat source (so far) has no Java 1.5 source syntax code in it. |
| 04:11.31 | swente | looking at the valve-source, i wondered if a couple of hardcoded "stringbuffer.append()" would be slower than a format-replacement |
| 04:12.44 | jasonb | Something you'll notice in the AccessLogValve source: some things look like they're coded weirdly, and when you finally figure out the reason, it comes down to performance. Try to change it and benchmark the diff and you'll find that your change runs slower than what is already there. :) |
| 04:12.54 | swente | ok :) |
| 04:13.09 | jasonb | But, VMs change, and hence the rules change.. so benchmarking is in order! |
| 04:13.54 | jasonb | Remy does just about everything on Windows, and the others on the list either do Windows or Linux. They all use the Sun JVM, so benchmarking another one won't get you anywhere with them. |
| 04:13.58 | swente | probably different platforms for the same VM might make a difference.. |
| 04:14.16 | jasonb | Remy will be the most critical of all patches, so if it runs slow on Windows, don't be surprised if he just says "this is too slow.. -1." |
| 04:14.23 | swente | hehehe :) |
| 04:14.39 | swente | well, i've to admit that i was surprised how fast tomcat is |
| 04:15.12 | jasonb | Yes, benchmarking some code on one OS with Sun's JVM gives you one set of numbers. Then, benchmarking it again on a different OS with the SAME JVM will give you a different set of numbers. |
| 04:15.35 | jasonb | The Tomcat committers (along with others) have literally spent years on making Tomcat run fast. |
| 04:15.54 | jasonb | Lots of benchmarking and refactoring for performance has been done already.. lots and lots of that. |
| 04:16.06 | swente | :) |
| 04:16.41 | jasonb | And, lots of would-be contributors modify the source to "fix" something, and they never benchmarked any of it before they submit it to the dev list, and they get shot down. |
| 04:16.45 | swente | i've just placed my first jsp-related site online, still relatively new to all of this |
| 04:17.39 | jasonb | Wow. Well, it's certainly worth spending your time on. Tomcat has been such a HUGE learning experience for me in all areas of computer science. That's why I've spent so many years on it. There's sooo much to learn there. |
| 04:18.06 | swente | mmh .. i also had a look at that j2ee-stuff, which impresses me too |
| 04:18.23 | jasonb | I've been writing and examining servlet containers (even ones other than Tomcat) since 1998. It's really been interesting. |
| 04:18.47 | jasonb | If you want to save yourself some time and frustration, stay away from EJB. |
| 04:19.14 | jasonb | Servlets, JSP, and servlet containers are great, though. |
| 04:19.43 | jasonb | Plus, there are many other facets of J2EE that are great. |
| 04:20.50 | swente | it's a wide field to explore |
| 04:21.37 | jasonb | It sure is. |
| 04:22.09 | jasonb | Just taking servlets.. that ends up being so wide that you can spend all your time on it and never run out of things to try & implement. :) |
| 04:22.46 | swente | i also discovered xml+xlst for me, recently |
| 04:25.08 | *** join/#tomcat dreivier (n=dreivier@port-212-202-43-109.dynamic.qsc.de) |
| 04:26.44 | jasonb | swente: That's another one. Heh. |
| 04:28.44 | *** join/#tomcat FreemaniaX (n=Freemani@pcp0011117215pcs.elkrdg01.md.comcast.net) |
| 06:16.01 | swente | jasonb: with a slightly modified (member visibility) AccessLogValve, my code reduced to a handful of lines:) |
| 06:16.10 | swente | jasonb: anyway, bedtime for me now |
| 07:34.51 | *** join/#tomcat Oleg (n=Oleg@dslb-084-059-218-023.pools.arcor-ip.net) |
| 09:05.48 | *** join/#tomcat omay (n=Oleg@dslb-084-059-216-086.pools.arcor-ip.net) |
| 09:28.57 | *** join/#tomcat valmont (n=chrishol@dsl092-043-004.lax1.dsl.speakeasy.net) |
| 10:14.16 | *** part/#tomcat Wulf4 (i=wulf@wulf.eu.org) |
| 10:15.26 | *** join/#tomcat linuxfreck (n=freck@p5492AADB.dip0.t-ipconnect.de) |
| 11:15.44 | *** join/#tomcat silverwolf (i=silverwo@83.93.43.159) |
| 11:47.22 | *** join/#tomcat vinse_ (n=vinse@ppp-71-139-33-236.dsl.snfc21.pacbell.net) |
| 12:10.10 | *** join/#tomcat geemark (n=geemark@port12.ds1-vey.adsl.cybercity.dk) |
| 12:50.49 | *** join/#tomcat silverwolf (i=silverwo@83.93.43.159) |
| 13:38.07 | swente | jasonb ? |
| 14:03.34 | *** join/#tomcat nicolasbolbol (i=nicooo@sav67-1-82-227-75-87.fbx.proxad.net) |
| 14:04.11 | nicolasbolbol | Hello |
| 14:05.51 | swente | 'lo nicolasbolbol |
| 14:05.58 | nicolasbolbol | :) |
| 14:06.21 | nicolasbolbol | is it necessary to restart tomcat when I change the content of a .java file ? |
| 14:06.55 | swente | nicolasbolbol: i guess you're using some ide? |
| 14:07.10 | nicolasbolbol | yes Eclipse |
| 14:07.25 | nicolasbolbol | and I begin crazy grrrrr |
| 14:07.35 | nicolasbolbol | everytime I get a 404 error and I don't know why |
| 14:07.50 | swente | nicolasbolbol: usually the context is reloaded when the content changes. i think it's a tomcat feature to watch the webroots. |
| 14:08.21 | nicolasbolbol | oki |
| 14:09.59 | nicolasbolbol | sheetttttttttttttttttttt |
| 14:10.06 | nicolasbolbol | can you help me I begin crazyyyyyy |
| 14:11.24 | swente | calm down, what's the problem? |
| 14:11.39 | swente | what do you mean by "everytime i get a 404" ? |
| 14:11.39 | nicolasbolbol | can I talk in private ? |
| 14:11.43 | swente | sure |
| 14:11.47 | nicolasbolbol | thanks |
| 14:47.08 | *** join/#tomcat FreemaniaX (n=Freemani@pcp0011117215pcs.elkrdg01.md.comcast.net) |
| 15:38.30 | *** join/#tomcat Snuff (n=tbel3393@CPE-149-167-219-246.sa.bigpond.net.au) |
| 16:39.19 | *** join/#tomcat lemniskate (n=lemniska@0127bhost185.starwoodbroadband.com) |
| 17:34.25 | *** join/#tomcat BayCanuck (n=chatzill@S01060080c80c77b5.cg.shawcable.net) |
| 18:24.38 | *** join/#tomcat FreemaniaX (n=Freemani@pcp0011117215pcs.elkrdg01.md.comcast.net) |
| 18:42.20 | *** join/#tomcat jaus (n=bk@62.57.85.110) |
| 18:42.31 | jaus | Hi, I use this on server.xml |
| 18:42.46 | jaus | port="127.0.0.1:8080" instead of port="8080" |
| 18:42.51 | jaus | it's a bad idea? |
| 18:43.08 | jaus | I use ajp connector from apache |
| 18:50.40 | *** join/#tomcat dabaR (n=dabaR@wnpgmb02dc1-184-86.dynamic.mts.net) |
| 18:51.22 | dabaR | Hi. How is tomcat installed on an Ubuntu computer? OPr in general onto a *nix compuiter from the web site? |
| 18:52.01 | dabaR | There are Unix daemon instructions in the documentation that I found, is that what I should follow? |
| 18:53.43 | dabaR | I haev to open screen this is ssh. brb. |
| 19:01.11 | *** join/#tomcat vinse_ (n=vinse_@ppp-71-139-33-236.dsl.snfc21.pacbell.net) |
| 19:04.22 | *** join/#tomcat dabaR (n=dabaR@wnpgmb02dc1-184-86.dynamic.mts.net) |
| 19:17.52 | *** join/#tomcat omay (n=Oleg@dslb-084-059-216-086.pools.arcor-ip.net) [NETSPLIT VICTIM] |
| 19:22.50 | jaus | if I use ajp connector from apache, i can disable Connector port=8080 http acces on server.xml on tomcat? |
| 19:24.27 | *** join/#tomcat vinse_ (n=vinse_@ppp-71-139-33-236.dsl.snfc21.pacbell.net) [NETSPLIT VICTIM] |
| 19:24.27 | *** join/#tomcat linuxfreck (n=freck@p5492AADB.dip0.t-ipconnect.de) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat swente (i=K1c3dhXn@hal.infinitumb.de) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat dabaR_ (n=dabaR@wnpgmb02dc1-184-86.dynamic.mts.net) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat puff (i=puff@66.45.34.102) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat fenris (n=fenris@brunhild.throughput.de) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat OneFix_Work (n=onefix@205.133.146.124) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat crackhead (i=mlm@blackbox.clarkk.net) [NETSPLIT VICTIM] |
| 19:24.28 | *** join/#tomcat strobedream (n=td@pcp05454736pcs.sandia01.nm.comcast.net) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat morgajel (n=morgajel@h-66-166-231-25.mclnva23.covad.net) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat manicman (n=manicman@unaffiliated/manicman) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat vito_ (n=vito@www.perspectivepartners.com) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat lilalinux (i=e-trolle@deepthroat.deswahnsinns.de) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat teite (i=shu@sun-shu.ision.net) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat pucko (n=andreas@h232n2fls31o809.telia.com) [NETSPLIT VICTIM] |
| 19:24.29 | *** join/#tomcat teddy (i=teddy@193.91.192.170) [NETSPLIT VICTIM] |
| 19:24.30 | *** join/#tomcat kousuke (n=kikuchi@221x251x19x21.ap221.ftth.ucom.ne.jp) [NETSPLIT VICTIM] |
| 19:24.30 | *** join/#tomcat lyken (n=janor@air665.startdedicated.com) [NETSPLIT VICTIM] |
| 19:24.30 | *** join/#tomcat Jara[zZ] (n=rimaraj@62.123.204.106) [NETSPLIT VICTIM] |
| 19:27.33 | *** join/#tomcat swente (i=K1c3dhXn@hal.infinitumb.de) [NETSPLIT VICTIM] |
| 19:27.33 | *** join/#tomcat puff (i=puff@66.45.34.102) [NETSPLIT VICTIM] |
| 19:27.33 | *** join/#tomcat fenris (n=fenris@brunhild.throughput.de) [NETSPLIT VICTIM] |
| 19:27.33 | *** join/#tomcat OneFix_Work (n=onefix@205.133.146.124) [NETSPLIT VICTIM] |
| 19:27.33 | *** join/#tomcat crackhead (i=mlm@blackbox.clarkk.net) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat strobedream (n=td@pcp05454736pcs.sandia01.nm.comcast.net) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat morgajel (n=morgajel@h-66-166-231-25.mclnva23.covad.net) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat manicman (n=manicman@unaffiliated/manicman) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat vito_ (n=vito@www.perspectivepartners.com) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat lilalinux (i=e-trolle@deepthroat.deswahnsinns.de) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat teite (i=shu@sun-shu.ision.net) [NETSPLIT VICTIM] |
| 19:27.34 | *** join/#tomcat Jara[zZ] (n=rimaraj@62.123.204.106) [NETSPLIT VICTIM] |
| 19:27.35 | *** join/#tomcat lyken (n=janor@air665.startdedicated.com) [NETSPLIT VICTIM] |
| 19:27.35 | *** join/#tomcat kousuke (n=kikuchi@221x251x19x21.ap221.ftth.ucom.ne.jp) [NETSPLIT VICTIM] |
| 19:27.35 | *** join/#tomcat teddy (i=teddy@193.91.192.170) [NETSPLIT VICTIM] |
| 19:27.35 | *** join/#tomcat pucko (n=andreas@h232n2fls31o809.telia.com) [NETSPLIT VICTIM] |
| 19:42.13 | dabaR | It being Tomcat. Is there a recommended directory for CATALINA_HOME? |
| 20:03.38 | dabaR | Does the .tar.gz for stable tomcat 4come with a manual? |
| 20:13.39 | jasonb | Not really. |
| 20:14.07 | jasonb | There are some docs, but I don't think I'd call them a manual. |
| 20:14.44 | jasonb | It's more like thin reference docs for terse explanations of a some of what you may need to know when administering Tomcat. |
| 20:14.57 | jasonb | (including configuration details) |
| 20:17.27 | jasonb | See the content in webapps/tomcat-docs/ |
| 20:26.57 | dabaR | wtheck is catalina anyhow? |
| 20:27.24 | *** join/#tomcat FreemaniaX (n=Freemani@pcp0011117215pcs.elkrdg01.md.comcast.net) |
| 20:29.40 | jasonb | Catalina was the name of the Tomcat 4.0 servlet container. |
| 20:29.54 | jasonb | It was reimplemented almost 100% from version 3.x. |
| 20:30.14 | jasonb | .. and thus got a new name.. until it was formallly dubbed Tomcat 4. |
| 20:30.25 | jasonb | The name lives on in the source code. |
| 20:31.08 | jasonb | Nowadays, Tomcat = Catalina (the servlet container), Jasper (the JSP implementation), Coyote (the HTTP server), and others. |
| 21:08.44 | FreemaniaX | tomcat provides a common db ppol that webapps can share via dbcp, right? |
| 21:08.47 | FreemaniaX | pool* |
| 21:09.48 | linuxfreck | FreemaniaX: right |
| 21:12.41 | *** join/#tomcat jwr (i=[EHnv45n@202.130.193.15) |
| 21:17.33 | dabaR | I started tomcat through ssh, and set up env variables(exported thenm too) using that ssh . Now once I log out, do I expect the tomcat to shut down, or to stay online? |
| 21:21.47 | linuxfreck | it should stay online |
| 21:29.54 | dabaR | Ill test |
| 21:33.20 | jaus | if I use ajp connector from apache, i can disable Connector port=8080 http acces on server.xml on tomcat? |
| 21:38.27 | jasonb | Sure. You may not be using it. |
| 21:38.51 | jaus | ok, thanks :) |
| 21:38.54 | jasonb | For debugging, sometimes being able to get in through 8080 as well is helpful. |
| 21:39.16 | jaus | but It is not mandatory nope? |
| 21:39.56 | jasonb | nope. :) |
| 21:40.58 | jaus | okey, thanks again |
| 21:42.42 | *** join/#tomcat jsisson (n=sissonj@216.9.46.68) |
| 21:51.10 | *** join/#tomcat bdudney (n=bdudney@c-24-9-189-43.hsd1.co.comcast.net) |
| 22:21.27 | FreemaniaX | is this the correct format for the start of a ROOT context.xml? |
| 22:21.28 | FreemaniaX | <Context path="/" docBase="ROOT" debug="99" reloadable="true"> |
| 22:24.10 | jasonb | Which version of Tomcat? :) |
| 22:25.02 | jasonb | Actually, I think it has to be <Context path="" ... |
| 22:29.32 | *** join/#tomcat Tonnerre (n=tonnerre@netbsd/pkgsrc/user/tonnerre) |
| 22:30.07 | FreemaniaX | 5.5.12 |
| 22:30.31 | FreemaniaX | path is an empty string? cool |
| 22:31.04 | FreemaniaX | is the docBase right? |
| 22:31.44 | Tonnerre | Excuse me, I have a question that the JSP examples don't seem to answer: is there any way to use arguments passed using GET or POST requests to the page in stuff like the book example? |
| 22:33.00 | Tonnerre | I want to be able to e.g. choose the book to be displayed via ISBN. Is that possible with the simpletag magic, or do I have to write standard JSP 1.2 pages for that? |
| 22:43.53 | jasonb | FreemaniaX: All I can really tell you about the docBase setting is: in my setup where I have a root context configured for TC 5.5.12, mine works and I have it configured like: <Context path="" reloadable="true"> (that's all, other than the </Context>) .. and that's in a file called ROOT.xml. |
| 22:47.10 | FreemaniaX | ah |
| 22:47.35 | FreemaniaX | I should really write quick tomcat tutorials |
| 22:47.55 | jasonb | FreemaniaX: Nah. people can just ask in here. :) |
| 22:47.55 | FreemaniaX | Things are not so obvious in the docs |
| 22:48.00 | FreemaniaX | lol |
| 22:48.08 | jasonb | FreemaniaX: What distro are you using? |
| 22:50.01 | FreemaniaX | osx :) |
| 22:50.13 | FreemaniaX | of java? tomcat? |
| 22:51.17 | vinse_ | hey jasonb |
| 22:51.22 | vinse_ | hows it |
| 22:51.59 | vinse_ | mind looking at something for me? you're familiar with ojb? |
| 22:52.07 | vinse_ | it's OT or i wouldnt bother asking ;) |
| 22:52.51 | FreemaniaX | osx 10.4 , tomcat 5.5.12, jdk 1.5 |
| 22:53.59 | Tonnerre | Does anyone have an answer for me? |
| 22:58.31 | swente | jasonb - are you there? |
| 23:06.52 | FreemaniaX | Tonnerre, you want to use request params in your jsp pages? |
| 23:07.36 | FreemaniaX | like: http://mysite/index.jsp?id=14 |
| 23:07.49 | Tonnerre | FreemaniaX, yes, I want like /book.jsp?isbn=0-... and have the isbn= passed to the simpletag handler |
| 23:08.09 | Tonnerre | Currently the thing already fetches its stuff from Postgres but I can't influence what it fetches |
| 23:09.38 | FreemaniaX | you can display the variable in a jsp 2.0 page with <b> this is the isbn ${param.isbn} </b> |
| 23:12.09 | Tonnerre | So getJspContext().getAttribute("param.isbn") in the Java code should do? |
| 23:12.23 | *** join/#tomcat dabaR_ (n=dabaR@wnpgmb02dc1-186-78.dynamic.mts.net) |
| 23:14.54 | FreemaniaX | why aren't you using EL Tonnerre ? |
| 23:15.20 | Tonnerre | EL? |
| 23:15.27 | Tonnerre | How would that work? |
| 23:31.19 | FreemaniaX | ohsix, you are talking about grabbing the param in your java class |
| 23:31.51 | FreemaniaX | Tonnerre, I meant |
| 23:32.06 | FreemaniaX | know how that is |
| 23:32.24 | jasonb | yeah |
| 23:32.29 | Tonnerre | FreemaniaX, I must get it into the Java class somehow |
| 23:33.27 | jasonb | swente: What's up? |
| 23:41.24 | swente | jasonb: ah ok, do you have a minute [accesslog-valve]? |
| 23:41.36 | jasonb | sure |
| 23:42.12 | swente | beside the fact that it's working [what i tinkered], i'm not sure if i'm on the wood-way |
| 23:42.49 | swente | basic problem was that i found no way to log to log4j [or jul] |
| 23:43.38 | swente | then i built that bridge to log4j, and thouht about creating a jul-bridge, too [it's tcp-based loging i want to try, too] |
| 23:44.16 | jasonb | swente: What do you mean you found no way to log to log4j or juli? |
| 23:44.34 | swente | jasonb: ehm .. does accesslogvalve support logging to jul or log4j? |
| 23:45.10 | jasonb | Not the way it is before any of your changes. But, being a component in the Catalina core, it *could* hook into the logging system that the rest of Catalina's objects use. |
| 23:45.40 | swente | and there we come to "global" configuration i guess? |
| 23:45.46 | jasonb | Yes. |
| 23:46.09 | jasonb | There is a global juli config file.. logging.properties. |
| 23:46.28 | swente | mmh ok, that was the point i was about .. the configuration .. my first attempt was kindof syslog-focusses, like having loghost+facility as valve-property |
| 23:47.03 | swente | (in fact i had a base-valve and an extended class, the 2nd bringing "property configuration") |
| 23:48.09 | jasonb | Well, if you're using juli, you should hook it into Tomcat's global logging.properties. |
| 23:48.14 | swente | so it would be probably best to keep as close to tomcat's current logging/configuration as possible |
| 23:48.16 | swente | mmh |
| 23:48.33 | jasonb | If you're using log4j, you can put your logging config file wherever, but that's more hacky. |
| 23:48.51 | swente | i guess that's why i wanted to have "my needed" options as "properties" :) |
| 23:49.45 | jasonb | IMHO basically any non-webapp config file should go in CATALINA_BASE/conf/. |
| 23:52.23 | swente | i've to admit i'm not very familiar to log4j and/or jul. the logging.properties, can one override them per-webapp? |
| 23:53.10 | swente | because, at least in my case i'd consider the access-log as webapp-part |
| 23:56.04 | jasonb | Actually, yes, you can configure logging per webapp. I think you just put a logging.properties file in WEB-INF/lib ?? Or was that just WEB-INF? |
| 23:56.36 | jasonb | But, keep in mind that valves are *not* part of webapps. |
| 23:58.16 | swente | oh, sure, i mean .. i need to configure them per-webapp, and would like to keep that configuration as part of them (as in WEB-INF/wherever). |
| 23:59.30 | jasonb | hmm, WEB-INF/classes is it. |