| 00:22.31 | *** join/#tomcat wsmoak (n=wsmoak@ip68-110-100-131.ph.ph.cox.net) |
| 00:33.12 | *** join/#tomcat kanai (n=kanai@124.32.114.226) |
| 00:46.12 | *** join/#tomcat bugfixer (n=bugfixer@42gis175.gulftel.com) |
| 00:48.10 | *** join/#tomcat pmcmahan (n=pmcmahan@adsl-065-005-198-110.sip.rdu.bellsouth.net) |
| 01:16.55 | *** join/#tomcat Marv|LG (n=Marv|TF@ip72-208-255-174.ph.ph.cox.net) |
| 01:19.25 | *** part/#tomcat kanai (n=kanai@124.32.114.226) |
| 01:24.29 | *** join/#tomcat jjchen (n=jjchen@58.246.26.222) |
| 01:25.14 | minerale | what port does ajp work on ? |
| 01:30.19 | nblracer | 8009 i think |
| 01:30.40 | nblracer | but im having troubles setting it up right now |
| 01:32.10 | pfn | ajp works on whatever port you configure it to |
| 01:33.19 | nblracer | 8009 is default for the deb i use |
| 01:33.21 | nblracer | d |
| 01:49.18 | *** join/#tomcat pfn (n=pfnguyen@hanhuy.com) |
| 01:49.19 | *** part/#tomcat locoo (n=locoo@207.249.166.83) |
| 02:23.57 | *** part/#tomcat pmcmahan (n=pmcmahan@adsl-065-005-198-110.sip.rdu.bellsouth.net) |
| 02:43.49 | *** join/#tomcat firemansam (n=iholsman@218-214-209-234.people.net.au) |
| 02:44.36 | firemansam | hey there. I was wondering if anyone knew of open source alternatives to tangosol |
| 03:05.56 | *** join/#tomcat ohsix (i=ohsix@bc111222.bendcable.com) [NETSPLIT VICTIM] |
| 03:10.59 | *** join/#tomcat minerale (i=achille@about/cooking/alfredo/Minerale) [NETSPLIT VICTIM] |
| 03:11.52 | *** join/#tomcat karstensrage (n=karstens@c-69-181-220-5.hsd1.ca.comcast.net) |
| 03:12.29 | karstensrage | the init method of a servlet only gets called once when the servlet is loaded, not for each connection, is that right? |
| 03:41.17 | evosh | karstensrage, Yes. |
| 03:41.28 | karstensrage | hmm |
| 03:42.06 | karstensrage | boy im having a bad problem |
| 03:42.12 | karstensrage | but I dont even know how to say it |
| 03:42.13 | evosh | Drugs? |
| 03:42.18 | karstensrage | no |
| 03:42.22 | evosh | Meth or Coke? |
| 03:42.30 | karstensrage | yoiu mind if I ramble a bit to try and explain it? |
| 03:42.55 | evosh | I am about to pass out so be my guest, entertain me. |
| 03:43.02 | karstensrage | i have this servlet |
| 03:43.27 | karstensrage | a load balancer is taking requests from a server and passing them to two instances of my servlet |
| 03:43.55 | karstensrage | the servlet creates "Jobs" and puts them in a LinkedBlockingQueue |
| 03:43.57 | evosh | Two instances of servlets on two Tomcat instances, right? |
| 03:44.00 | karstensrage | yes |
| 03:44.04 | evosh | Ok. |
| 03:44.17 | karstensrage | on another connector (8081) |
| 03:44.57 | karstensrage | there are 60 machines that make requests of the same servlet (on the other port) to get the "Jobs" from the LinkedBlockingQueue |
| 03:45.11 | karstensrage | each of the 60 machines runs 4 threads that does this |
| 03:45.18 | karstensrage | so 240 requests coming in |
| 03:45.38 | karstensrage | those machines, have a method to hit one or the other servlet |
| 03:45.44 | karstensrage | so half hit one, half hit the other |
| 03:45.58 | karstensrage | everything works perfectly in steady state |
| 03:46.17 | karstensrage | the problem occurs when you take one of the tomcat's down |
| 03:46.43 | karstensrage | the load balancer sees the servlet down and starts sending all requests to the other tomcat |
| 03:47.07 | karstensrage | the 30 machines that were going to that tomcat on 8081... move over to the other tomcat |
| 03:47.12 | karstensrage | so far so good |
| 03:47.31 | karstensrage | now when you add the tomcat that went down back in, bring it up |
| 03:48.05 | karstensrage | it gets like 8 or 10 requests from the load balancer but the 30 machines take a lot longer to migrate back to the new tomcat |
| 03:48.27 | karstensrage | if I can make a manual request (through a browser) I can get the servlet to kickstart and it runs fine |
| 03:48.59 | karstensrage | but if I dont make it fast enough, ther requests coming in saturate the servlet and it goes down.. doest respond anymore |
| 03:49.21 | karstensrage | in the status page you can see that it gets all the way to 200 busy threads and then just hangs |
| 03:49.32 | karstensrage | Im not sure what to even try |
| 03:50.19 | karstensrage | i moved the Queue to a BlockingQueue (originally it was a just a LinkedList and they tried to do the home grown synchronization on it) |
| 03:51.00 | karstensrage | now one interesting thing is that we migrated to JDK 1.5_09 |
| 03:51.04 | karstensrage | and TC 5.5.20 |
| 03:51.12 | karstensrage | from JDK 1.4.2 |
| 03:51.16 | karstensrage | and TC 4.4 |
| 03:51.34 | karstensrage | and the old version seems to be more stable in this particular scenario |
| 03:52.20 | karstensrage | we noticed another significant difference with the new setup, we had to explicitly close all the streams that were opened, to get TC to release its connections |
| 03:53.28 | evosh | I only know ABC of Tomcat so much so for clustering (no clue how to even implement that). My first naive question would be, what a mess! |
| 03:53.47 | evosh | You know, I've also noticed Tomcat 5.5 hold the connections for some time and I can't figure out why. |
| 03:54.08 | karstensrage | close your input and output streams after you are done with them |
| 03:54.27 | evosh | So let me clarify this, the newly revived Tomcat only comes back to life after manual request via browser? |
| 03:54.39 | karstensrage | onlhy if I do it fast enough |
| 03:54.55 | karstensrage | if I can get a manual request in before it gets saturated, it will kick start it |
| 03:54.57 | evosh | And only after that it gets flooded with request? So what about the second Tomcat instance, does that one also get saturated? |
| 03:55.19 | karstensrage | but its hard to get it in since the load balancer keeps sending requests to it once it sees that it came back up |
| 03:55.31 | evosh | And how are your "job" machines locate the Tomcat machines and how requests are handled? |
| 03:55.48 | karstensrage | no the second one keeps running |
| 03:56.11 | karstensrage | the job machines have the ip addresses of the tomcat machines |
| 03:56.33 | karstensrage | the requests are handled by the servlet |
| 03:56.38 | karstensrage | not sure what you mean by that? |
| 03:56.41 | evosh | But how are they notified that the dead instance has come back online/ |
| 03:57.27 | karstensrage | they arent, they have the notion of the backup (the "other" one) and they try that one for a configured amount of time, usually one minute, after that they try the original one again |
| 03:58.00 | karstensrage | but it doesnt seem to have anything with those |
| 03:58.22 | evosh | Ok, because the Tomcat instances notify the "job" machines as whether they are online or not. |
| 03:58.26 | karstensrage | because if I get that manual request in, everything is fine, and it continues to be fine when those 30 machines come back |
| 03:58.52 | evosh | Well, the problem I would assume lies in the load balancer. |
| 03:58.52 | karstensrage | well only by responding to the request |
| 03:59.15 | karstensrage | how do you figure? |
| 03:59.38 | evosh | Why would the load balancer saturate the newly revived instance? |
| 03:59.51 | karstensrage | it sees it as up, and starts sending requests to it |
| 03:59.59 | evosh | I mean, it's not the job of the machines (60) to figure how which instance to hit, right? |
| 04:00.35 | evosh | They just pop the queue for a job so they shouldn't care about which instance to hit. |
| 04:00.41 | karstensrage | yeah |
| 04:00.48 | evosh | But the job of the load balancer is to balance these requests. |
| 04:01.27 | *** join/#tomcat vinse (n=vinse@208.253.223.146) [NETSPLIT VICTIM] |
| 04:01.27 | evosh | Frankly, I have never worked with the load balancer (software or hardware) so I probably not be a good candidate to answer your question. |
| 04:01.27 | karstensrage | lol |
| 04:01.35 | evosh | But a lot of people come here ask about session handling of load balancer. |
| 04:01.36 | karstensrage | me neither, I am just fixing someone else code |
| 04:02.10 | evosh | Where is that someone else? |
| 04:02.38 | karstensrage | well hes around, but he kinda washes his hands of it since we upgraded to 1.5 and 5.5.20 |
| 04:03.13 | karstensrage | I wish I knew what the differences were with the underlying handling of connections with 5.5 |
| 04:03.19 | evosh | So out of curiosity, what load balancer are you utilizing? |
| 04:03.22 | evosh | damn spelling. |
| 04:03.47 | karstensrage | honestly I dont know, hold on, maybe someone is online that knows |
| 04:05.48 | karstensrage | he might respond, i dont nkow |
| 04:05.51 | evosh | A few people knows more about connectors but as you can see, it's a slow moving channel; you gotta wait a while (sometimes days). |
| 04:05.52 | karstensrage | know |
| 04:06.00 | karstensrage | hehe |
| 04:06.04 | karstensrage | do they scroll back? |
| 04:06.14 | karstensrage | or do I have to publish that story :) |
| 04:06.36 | karstensrage | a citrix netscaler |
| 04:06.38 | evosh | But in any case, your situation has too many variables invovled. It'll be hard for someone, if any, to diagnose the problem. |
| 04:06.43 | karstensrage | thats the load balancer |
| 04:06.55 | evosh | Ok, so it's a hardware. |
| 04:07.07 | karstensrage | i guess so |
| 04:07.20 | evosh | If you see "jasonb", hollar at him -- he IS the man. |
| 04:07.28 | karstensrage | yeah |
| 04:07.37 | evosh | pfn is very knowledgable but haven't see him today. |
| 04:07.58 | karstensrage | well thanks for listening |
| 04:08.20 | evosh | I enjoy learning something new any time. |
| 04:08.33 | evosh | Even if it means through someone else's problem. |
| 04:08.48 | evosh | Sorry couldn't help... Like I ever be able to but doesn't hurt to try. |
| 04:09.22 | karstensrage | well have you heard of the programming methodology called "talking to the duck" |
| 04:09.38 | evosh | Hah, ya. |
| 04:09.44 | karstensrage | you are my duck :) |
| 04:10.06 | karstensrage | just saying it makes me think about it and there are some things I can try tomorrow |
| 04:10.38 | evosh | Hah, ok and since this the last topic gone through my head tonight, I'll probably be dreaming about it. |
| 04:10.40 | evosh | Scary... |
| 04:10.41 | karstensrage | look at what a manual request does and maybe throw the jobs away if there are no machines connecting |
| 04:11.02 | karstensrage | if you come up with anything, Ill leave the IRC on and you can tell me |
| 04:11.10 | karstensrage | the IRC client that is |
| 04:11.12 | evosh | Ok, for sure. |
| 04:11.17 | karstensrage | have good night |
| 04:11.29 | evosh | 'night |
| 04:15.19 | *** join/#tomcat firemansam (n=iholsman@218-214-209-234.people.net.au) |
| 04:31.28 | *** join/#tomcat kanai_ (n=kanai@124.32.114.226) |
| 04:34.58 | *** join/#tomcat doorman (n=n@cpe-76-171-7-115.socal.res.rr.com) |
| 04:39.50 | doorman | hi guys, i'm sorry if this is a dumb question... i've installed tomcat 6.0.13 on my school account, running SunOS 5.9 on the server, winxp on the client, java-1.5.0... now i need to host a website on the tomcat server - where should i place the html files? ie in which directory??? |
| 04:54.15 | *** join/#tomcat jasonb (i=noneoyer@adsl-66-124-73-250.dsl.sntc01.pacbell.net) |
| 04:55.49 | doorman | hi guys, i'm sorry if this is a dumb question... i've installed tomcat 6.0.13 on my school account, running SunOS 5.9 on the server, winxp on the client, java-1.5.0... now i need to host a website on the tomcat server - where should i place the html files? ie in which directory??? |
| 04:58.07 | jasonb | doorman: You either need to read a book about Java servlets and Java webapps, or you need to have a look at the Java Servlet Specification. |
| 04:59.27 | jasonb | doorman: First, you might want to look at this: http://tomcat.apache.org/tomcat-6.0-doc/appdev/index.html |
| 05:01.08 | doorman | good heavens |
| 05:01.36 | doorman | ahh yes i was reading that page right now |
| 05:02.32 | jasonb | doorman: It may sound scary, but it's not. Just have a look. |
| 05:02.54 | doorman | okay |
| 05:03.15 | jasonb | doorman: Basically, you are wanting to start a new Servlet web application. |
| 05:03.34 | doorman | and produce a war file? |
| 05:03.45 | jasonb | With Tomcat, you do not have to create a war file. |
| 05:03.54 | pfn | well, in tomcat, you can just use the webapps/ROOT directory to host everything... |
| 05:04.01 | jasonb | Tomcat allows the webapps to be deployed as a directory of web content. |
| 05:04.08 | pfn | really? |
| 05:04.14 | jasonb | Of course. |
| 05:04.28 | jasonb | In fact, I highly suggest and prefer unpacked webapp dirs. |
| 05:04.50 | jasonb | WAR files are an unnecessary, unhelpful distraction, in just about all cases. |
| 05:05.07 | doorman | oh ok |
| 05:05.23 | doorman | yeh i uploaded it to the webapps/ROOT directory - works! |
| 05:05.47 | doorman | jasonb - what was that about servlets? |
| 05:05.54 | jasonb | doorman: In that case, you're adding to the ROOT webapp. |
| 05:06.36 | jasonb | doorman: "Servlets" is the main standard that Tomcat implements. Tomcat really only serves Servlet-compliant web applications. |
| 05:06.51 | doorman | ohh |
| 05:06.52 | jasonb | doorman: And, Tomcat treats everything it serves as being part of a web application. |
| 05:07.13 | jasonb | doorman: And, Tomcat assumes that it will be serving multiple webapps. |
| 05:07.47 | doorman | i see |
| 05:08.35 | jasonb | So, they're "servlet webapps". |
| 05:09.29 | doorman | okay |
| 05:09.39 | doorman | man i really need to learn java properly |
| 05:10.11 | jasonb | Just take it one day at a time. :) |
| 05:10.31 | doorman | any particular books you'd suggest for a newbie? |
| 05:10.39 | jasonb | Of course. :) |
| 05:10.52 | jasonb | First, you're in the right place to ask questions. |
| 05:10.52 | doorman | which? |
| 05:11.05 | jasonb | You can get lots of good answers in here. |
| 05:11.18 | doorman | lol yeh i found this irc channel at tomcat.apache.org |
| 05:11.23 | jasonb | Plus, other Java related channels. |
| 05:11.28 | jasonb | There's also ##java. |
| 05:11.49 | jasonb | (although some of the guys in there act like eliteists.) |
| 05:12.05 | jasonb | (even though they're actually lame) |
| 05:12.13 | doorman | is there someone called koalaman there? |
| 05:12.22 | jasonb | Hmm, probably not in here. |
| 05:12.27 | doorman | different server maybe |
| 05:12.41 | doorman | so is there any good book in particular? |
| 05:12.45 | doorman | like headfirst java? |
| 05:12.51 | jasonb | What would you want to learn? |
| 05:13.13 | doorman | well java programming for starters |
| 05:13.23 | doorman | i've been stuck in C and sockets for too long |
| 05:13.33 | jasonb | One thing I suggest for that is http://java.sun.com/tutorial |
| 05:14.21 | jasonb | The Sun Java Tutorial is also printed as a book, but you might as well read it online. |
| 05:14.27 | jasonb | (it's more up to date online) |
| 05:14.36 | doorman | right |
| 05:15.18 | jasonb | There's also the Java In a Nutshell book .. (latest edition) |
| 05:15.49 | jasonb | Learn the basics, then get the O'Reilly Java Threads book! |
| 05:16.27 | doorman | and the basics wld be doable from java.sun.com/tutorial right? |
| 05:16.45 | jasonb | Well, yes. There is a lot of very good, detailed content in the Sun tutorial. |
| 05:16.58 | jasonb | And, it's free to read it. |
| 05:17.31 | doorman | cool |
| 05:42.39 | *** join/#tomcat firemansam (n=iholsman@218-214-209-234.people.net.au) |
| 06:05.07 | *** join/#tomcat ibot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 06:05.08 | *** topic/#tomcat is Stable versions: 6.0.13, 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. |
| 06:05.43 | jasonb | pfn: Some of the same old problems exist in F7 that I was hoping would get fixed, but oh well, it does look nicer, overall, than FC6. |
| 06:06.15 | pfn | what problems in fc6? |
| 06:06.17 | jasonb | pfn: One thing though, F7 does not come with compat-gcc-3.x anymore, nor compat libs, which means lots of binaries won't run on it, sadly. |
| 06:06.42 | jasonb | Well, like FC6 and lower came with broken GNU Java, and there wasn't anything you could do to make it NOT install that.. same with F7. |
| 06:06.46 | wltjr | jasonb: people and their silly binaries :) |
| 06:06.59 | pfn | oh, I don't have gcj... |
| 06:07.05 | pfn | I did a yum upgrade, though |
| 06:07.06 | jasonb | Also, F7 still comes with Tomcat 5.5, not 6.x.. bad. |
| 06:07.35 | jasonb | pfn: Do a "which java" in a shell where you have not set JAVA_HOME to a hand-installed JDK. |
| 06:08.00 | jasonb | pfn: I even unchecked the box that was the package group for "Java", and it installed it anyway! |
| 06:08.02 | jasonb | (today) |
| 06:08.13 | pfn | [pfnguyen@ares ~]$ which java |
| 06:08.13 | pfn | /usr/bin/java |
| 06:08.13 | pfn | [pfnguyen@ares ~]$ ls -l /usr/bin/java |
| 06:08.13 | pfn | lrwxrwxrwx 1 root root 26 2007-05-07 14:40 /usr/bin/java -> /usr/java/default/bin/java |
| 06:08.16 | pfn | [pfnguyen@ares ~]$ ls -l /usr/java/default/bin/java |
| 06:08.19 | pfn | -rwxr-xr-x 1 root root 47116 2007-03-14 03:58 /usr/java/default/bin/java |
| 06:08.24 | pfn | sorry, brief scroll |
| 06:08.28 | pfn | but since we're low traffic ;-) |
| 06:08.28 | pfn | [pfnguyen@ares ~]$ java -version |
| 06:08.28 | pfn | java version "1.6.0_01" |
| 06:08.32 | jasonb | wltjr: Yeah, well, there are lots of binaries that come only as binaries in the real world.. like it or not. |
| 06:09.03 | jasonb | pfn: You must have set alternatives already. |
| 06:09.11 | pfn | jasonb no, I did a yum upgrade... |
| 06:09.15 | jasonb | pfn: Did you upgrade an existing installation of an older fedora? |
| 06:09.24 | pfn | fc6->f7 |
| 06:09.32 | jasonb | pfn: How did you do that? What command(s) exactly? |
| 06:09.35 | wltjr | not so much in my real world ;) but yes binaries are out there, only ones I use primarily are closed source |
| 06:09.37 | pfn | so it doesn't install anything except for whatever is already on the machine |
| 06:09.43 | pfn | http://fedoraproject.org/wiki/YumUpgradeFaq |
| 06:09.49 | pfn | it's an "unsupported" upgrade mechanism |
| 06:09.49 | jasonb | I've never attempted that before, and probably won't, but I'm interested in how people do it. :) |
| 06:09.57 | pfn | but it's damn convenient |
| 06:10.01 | pfn | no need to download a giant iso |
| 06:10.01 | wltjr | jasonb: we got java compiling :) |
| 06:10.03 | pfn | or burn a disk |
| 06:10.07 | pfn | or reboot even |
| 06:10.10 | pfn | (except when you're done) |
| 06:10.20 | wltjr | OpenJDK 1.7.0.0_alpha13 ;) |
| 06:10.23 | pfn | and on top of that, you don't get all the unnecessary packages you don't want |
| 06:10.33 | jasonb | wltjr: I heard. Are you using it already? |
| 06:10.54 | jasonb | pfn: You really should reboot when you're done with that yum command! |
| 06:10.57 | wltjr | yes, haven't switched just yet on my dev server, but on all workstations, mainly for nb and development |
| 06:11.08 | wltjr | let me go switch to it now on my dev server :) |
| 06:11.14 | pfn | jasonb well, I reboot on new kernels, and yes, there's a new kernel |
| 06:11.17 | *** part/#tomcat firemansam (n=iholsman@218-214-209-234.people.net.au) |
| 06:11.25 | wltjr | takes a bit to compile though :( |
| 06:11.38 | jasonb | wltjr: That's very cool.. can you do a java -version and paste the output in here? |
| 06:11.44 | wltjr | hmm, looks as if I already did on my dev server :) |
| 06:11.58 | pfn | I wonder if my @Property annotation still works on java1.7 |
| 06:11.58 | pfn | heh |
| 06:12.20 | wltjr | wlt@wlt ~ $ java-config -v |
| 06:12.20 | wltjr | openjdk version "1.7.0-internal" |
| 06:12.21 | wltjr | OpenJDK Runtime Environment (build 1.7.0-internal-root_11_may_2007_12_00-b00) |
| 06:12.25 | wltjr | OpenJDK 64-Bit Server VM (build 1.7.0-internal-root_11_may_2007_12_00-b00, mixed mode) |
| 06:12.36 | wltjr | I got 13 on my x86 workstatoins |
| 06:13.01 | wltjr | OpenJDK Runtime Environment (build 1.7.0-internal-root_30_may_2007_21_18-b00) |
| 06:13.01 | wltjr | OpenJDK Client VM (build 1.7.0-internal-root_30_may_2007_21_18-b00, mixed mode) |
| 06:13.02 | jasonb | Upgrading Fedora Using Yum Frequently Asked Questions: Don't! :) |
| 06:13.33 | jasonb | wltjr: That's sweet! |
| 06:13.57 | wltjr | sadly no effort on my behalf but others have been hard at it, already providing bug reports and patches to Sun, and etc |
| 06:13.58 | jasonb | wltjr: Why isn't it "java -version" though? |
| 06:13.59 | pfn | jasonb yeah, they say don't, but it works |
| 06:14.14 | wltjr | find exploits, using system libraries instead of bundled ones |
| 06:14.32 | wltjr | oh whoops habit |
| 06:14.35 | jasonb | pfn: Yes, it works. I know enough about how RPM works to know the things that might break, though, so I agree with them! |
| 06:14.35 | wltjr | should be same thing |
| 06:14.45 | wltjr | wlt@wlt ~ $ java -version |
| 06:14.45 | wltjr | openjdk version "1.7.0-internal" |
| 06:14.45 | wltjr | OpenJDK Runtime Environment (build 1.7.0-internal-root_11_may_2007_12_00-b00) |
| 06:14.45 | wltjr | OpenJDK 64-Bit Server VM (build 1.7.0-internal-root_11_may_2007_12_00-b00, mixed mode |
| 06:14.58 | jasonb | wltjr: That's very cool! |
| 06:15.01 | wltjr | java-config is just our well, java config and management tool :) |
| 06:15.11 | wltjr | it's part of what every other OS in the world lacks ;) |
| 06:15.12 | lyken | oo |
| 06:15.16 | lyken | is that OOjava ? |
| 06:15.35 | wltjr | http://rafb.net/p/O4B0mW32.html |
| 06:15.36 | pfn | jasonb I know how things break with rpm all the time, especially when I used checkinstall and did an upgrade from fc4 -> fc6 |
| 06:15.40 | pfn | that was an annoyance |
| 06:16.06 | wltjr | I can switch very easily java-config -S 1-6, or java-config -S sun-jdk-1.5 |
| 06:16.28 | wltjr | you have a system vm, user vm, build vm :), choice of compilers |
| 06:16.37 | wltjr | java-config also can provide jars and build a cp for you :) |
| 06:17.46 | jasonb | wltjr: 1.7.0-internal... internal to what? I'd say it's pretty external! :) |
| 06:17.46 | wltjr | http://rafb.net/p/S6Vkqa62.html |
| 06:18.16 | wltjr | jasonb: likely their Sun's tagging, they are still in the transition period of opening up their process to the outside world :) |
| 06:18.31 | wltjr | jasonb: kinda like how freaking tc devs aren't tagging their sources anymore :( |
| 06:18.33 | jasonb | I guess they are. |
| 06:18.41 | jasonb | heh |
| 06:18.44 | pfn | why should their source be tagged? |
| 06:18.47 | jasonb | I saw the messages about that. |
| 06:18.48 | pfn | that's just a cvs-ism |
| 06:18.53 | wltjr | 5.5.23 prop.defaults file has 5.5.20, and 6.0.x has always had 6.0-snapshot :( |
| 06:19.07 | pfn | what are you gonna tag? |
| 06:19.17 | pfn | -snapshot-rc1, -rc2, rc3? |
| 06:19.19 | wltjr | sure, but I liked the suggestion about doing it via their dist.xml or as part of release process so does not require a commit or etc |
| 06:19.22 | pfn | only if it's an official release |
| 06:19.30 | wltjr | but most tend to tag their sources in cvs/svn, but won't impose that |
| 06:19.37 | pfn | well, a tag requires a commit |
| 06:19.49 | pfn | not like that's a pain point or anything |
| 06:19.58 | wltjr | well it's their sources, and the sources are externally versioned, just not internally :) |
| 06:20.14 | pfn | well, if you want a snapshot by day, you can get that without a tag |
| 06:20.15 | jasonb | I agree that tagging is mostly optional with subversion, but at least there should be something in the repo about which rev a release was made from. |
| 06:20.23 | wltjr | I am not talking about that |
| 06:20.28 | pfn | you mean there's no tag for 6.0.13? |
| 06:20.31 | pfn | really? |
| 06:20.33 | pfn | ok, now that's fucked up |
| 06:20.51 | pfn | except it's perl |
| 06:20.55 | lyken | svn ... |
| 06:20.55 | wltjr | if I download this http://apache.cs.utah.edu/tomcat/tomcat-6/v6.0.13/src/apache-tomcat-6.0.13-src.tar.gz |
| 06:20.55 | jasonb | svn? |
| 06:20.59 | jasonb | :) |
| 06:21.11 | wltjr | and unpack and type ant, and just invoke default build, it will spit out 6.0-snapshot :( not 6.0.13 |
| 06:21.22 | wltjr | 5.5.23 sources spit out 5.5.20 :( |
| 06:21.42 | pfn | svk |
| 06:21.50 | pfn | svk is a distributed "extension" to svn |
| 06:21.59 | wltjr | it's just funny to see how open source projects neglect their source releases :) |
| 06:22.05 | pfn | although it also supports cvs and p4 as central repositories |
| 06:22.13 | jasonb | http://svn.apache.org/repos/asf/tomcat/tc6.0.x/tags/ |
| 06:22.18 | pfn | wltjr, makes sense to me... |
| 06:22.41 | wltjr | pretty sure it was always versioned in the past, not sure what's going on |
| 06:22.44 | pfn | it just means they don't have a script that generates the version based on code dropped |
| 06:23.04 | pfn | interesting, though |
| 06:23.07 | wltjr | jasonb: you hear anything about grizzly? |
| 06:23.09 | pfn | there's no branch for 6.0-snapshot |
| 06:22.51 | jasonb | wltjr: Of course. |
| 06:22.54 | wltjr | I believe that's the tc equivalent in glassfish no? |
| 06:23.15 | wltjr | jasonb: slowly looking to take bites out of and package gf, oh the nightmare |
| 06:23.18 | pfn | svk is cool, working copies are 1/3 the size of an svn wc |
| 06:23.28 | jasonb | wltjr: I attended Francois Arcand's Grizzly connector session at JavaOne, met him, got a Grizzly T-shirt even! |
| 06:23.30 | pfn | and no stupid .svn files |
| 06:23.56 | wltjr | jasonb: cool |
| 06:24.13 | wltjr | jasonb: next year I will be at JavaOne for sure, I got a way in without shelling out the $ |
| 06:24.17 | jasonb | Grizzly is just a NIO framework with a web server built on top of it.. so it's like the NIO connector in Tomcat. |
| 06:24.23 | wltjr | I will help man/staff a Gentoo booth :) |
| 06:24.41 | jasonb | lyken was at the gentoo booth this year.. he should tell you about that. :) |
| 06:24.50 | wltjr | jasonb: yeah it's quite confusing since they call it a connector, not app/web server or etc |
| 06:25.07 | wltjr | really so there was one and it was staffed? news to me :) |
| 06:25.12 | lyken | err wot ? |
| 06:25.19 | lyken | crack munkey |
| 06:25.27 | lyken | i was manning the postgres booth |
| 06:25.28 | jasonb | pfn: I think I've looked at the svk page before, and came away confused about what it was. Do you have a link? |
| 06:25.32 | lyken | and talking smack at the debian guys |
| 06:25.41 | pfn | jasonb svk.bestpractical.com or something |
| 06:25.59 | pfn | jasonb just think of it this way, you mirror the repo locally, and you can have local branches |
| 06:26.12 | jasonb | lyken: Oh, debian.. Yeah, sometimes I get debian and gentoo confused. |
| 06:26.14 | pfn | jasonb imagine you're on a plane and coding (or a coffee shop w/o net access, a park, etc.) |
| 06:26.21 | pfn | now, what do you do if you want to checkpoint your changes? |
| 06:26.42 | pfn | or, say you use an OSS software that you build from source |
| 06:26.53 | jasonb | pfn: That does sound cool. |
| 06:26.55 | wltjr | jasonb: interesting distros to mix, quite different philosophies and approaches |
| 06:27.01 | pfn | what do you do if you want to make changes but you're not a committer on the project |
| 06:27.11 | pfn | or your patches are not acceptable/accepted? |
| 06:27.16 | pfn | or you don't want to submit your patches? |
| 06:27.26 | pfn | jasonb that's what svk gives you |
| 06:27.48 | pfn | svk gives you a completely independent version control repository |
| 06:27.49 | jasonb | wltjr: Yes, I know. But, they're alike to me in that they both seem fundamentally broken. :) But, I'll certainly give gentoo the credit for being not as broken as debian! |
| 06:27.56 | wltjr | interesting, have to look into svk |
| 06:28.05 | pfn | and allows you to push/pull changes from the "central" repository at any time |
| 06:28.17 | pfn | you can also make it play nice with standard svn clients... |
| 06:28.23 | pfn | it also improves merging |
| 06:28.34 | pfn | since it provides merge tracking, something vanilla svn lacks |
| 06:28.42 | wltjr | jasonb: well really we should not be broken, have come quite a long way, now 5.5.x is crap, but 6.0.x should be pretty solid |
| 06:29.31 | wltjr | I say 5.5.x is crap mainly because of it's deps and etc, but should be solid and usable on Gentoo |
| 06:29.51 | jasonb | wltjr: Gentoo and Blastwave are the only distros that I know of that right now have Tomcat 6.x packages, so that's certainly something. |
| 06:30.03 | wltjr | but I need to do much more, implement vhosts for Tomcat, come up with a way to package webapps, so they can be used by more than one vhost, and etc |
| 06:30.21 | wltjr | jasonb: our stuff get's lots of luv by me |
| 06:30.36 | wltjr | tc stuff is in tree same day it's released, and usually when I have time I will do dev snapshots adn etc |
| 06:30.48 | wltjr | http://viewcvs.gentoo.org/viewcvs.py/gentoo-x86/www-servers/tomcat/?hideattic=0 |
| 06:31.02 | wltjr | start from like 5.5.x on, I did a bit with 5.0.28, nothing with anything before |
| 06:31.03 | jasonb | wltjr: That's something the other distros lack, mainly because a company has a stranglehold and doesn't allow the community to give the packages the love they deserve... all except for Debian, which is just junk. |
| 06:31.22 | *** join/#tomcat fix_ (n=fix@cable-87-244-191-178.upc.chello.be) |
| 06:31.46 | wltjr | been busting my hump lately to resolve most all open bugs, had like 17 or so in the past 1.5 months :( but mostly gentoo bs stuff |
| 06:32.38 | wltjr | jasonb: 6.0.x would be stable, but .10 had a nio leak and I did not feel comfortable stabilizing it with a known bug upstream, and new release with fixes on the way |
| 06:32.57 | wltjr | short of any bugs or changes, 6.0.13 will be in ~30 days |
| 06:33.16 | jasonb | wltjr: Your page looks similar to mine at http://www.webdroid.org:8080/archives/tomcat-package only mine is less up to date, and is just a dir full of packages. |
| 06:33.18 | wltjr | jasonb: we have tomcat-native and all :) |
| 06:33.36 | wltjr | http://packages.gentoo.org/search/?sstring=tomcat |
| 06:34.10 | wltjr | and servlet-api is ripped out and packaged in it's own, since most any package using a serlvet-api.jar get's it from tc :) how, I don't want to know :) |
| 06:34.45 | jasonb | wltjr: That's cool. I wrote up detailed instructions about how to build and run the APR stuff for all other distros in the Tomcat book, so I'm sure it'll get more use. But, according to my benchmarks, it's not worth spending time on. |
| 06:35.18 | wltjr | jasonb: yeah and most run tc behind apache or etc, so unless they run stand alone with ssl really not much benefit as I understand it |
| 06:36.03 | jasonb | wltjr: I'm hoping the text in our book will pursuade a large percent of the Tomcat population to stop running Tomcat behind httpd. It's worded quite carefully and convincingly. |
| 06:36.10 | wltjr | jasonb: I have had a decent amount of interest on Gentoo with it, although not much feedback on performance or etc, not sure if they are using it, or seeing the message in the logs and think they are missing out on something special :) |
| 06:36.29 | wltjr | jasonb: I would not if I did not have sites and etc that aren't java |
| 06:36.51 | wltjr | not sure I want to run a bunch of static sites, and perl stuff via tomcat :) |
| 06:37.01 | wltjr | much less like mod_python or etc |
| 06:37.04 | jasonb | wltjr: When you read my text, some day, :) you'll realize there are options open to you that you never thought of. |
| 06:37.31 | wltjr | jasonb: sure, I see php can run in tc, and cf runs on java, so I am sure there is lots more out there |
| 06:37.45 | wltjr | oldschool chilisoft asp was also running on java |
| 06:38.07 | jasonb | wltjr: Running everything non-java inside of Tomcat is tedious and inelegant, so I don't expect people to do that. But, there are alternatives. |
| 06:38.16 | wltjr | jasonb: it's just the separation I like, plus ssl and java can be a pita, if I have to deal with keystore or etc |
| 06:38.33 | jasonb | wltjr: Not anymore. |
| 06:38.38 | wltjr | jasonb: sure for clients, that run just their own apps, with tc and apache for once site, I might lose apache |
| 06:38.51 | jasonb | wltjr: You don't have to do anything with keytool if you don't want to. |
| 06:38.56 | wltjr | in fact, I might consider that for one client in particular, I don't believe they have any need for apache |
| 06:39.29 | wltjr | jasonb: not required for certs? I have to use it to import certs for java to use on the desktop at times |
| 06:39.38 | jasonb | wltjr: You're not thinking far enough outside the box. |
| 06:39.57 | wltjr | well if tc can have certs configured like apache or etc, that's great |
| 06:40.07 | jasonb | wltjr: You can set up Tomcat with HTTPS without ever using keytool. |
| 06:40.16 | wltjr | really never explored ssl much with tc, I believe I messed with it in 3.x or 4.x days, but that's YEARS ago for me :) |
| 06:40.16 | jasonb | wltjr: It can. |
| 06:40.35 | wltjr | I can't even recall when I last ran 4.x |
| 06:41.07 | jasonb | 4.x should be looong gone, looong loong gone! |
| 06:41.50 | wltjr | seems many still use it in production, since 5.x had many bugs just now fixed in like 5.5.23 or etc |
| 06:42.07 | jasonb | People still use JServ in production too. |
| 06:42.12 | wltjr | one dude in particular said on -dev ml that he had no choice |
| 06:42.15 | jasonb | Not smart people, but people. |
| 06:42.28 | wltjr | well his was due to a know bug that was fixed in cvs :) |
| 06:42.58 | wltjr | plus they do allot of jumping |
| 06:43.08 | jasonb | jumping? |
| 06:43.13 | wltjr | 5.5.17 -> 5.5.20 -> 5.5.23 |
| 06:43.23 | jasonb | why? |
| 06:43.24 | wltjr | 6.0.10 -> 6.0.13 :) |
| 06:44.03 | jasonb | I think that doesn't exist anywhere else.. so far. |
| 06:44.08 | wltjr | usually release versions increment consistently without gaps, all total moot issues, just oddities with tc that I don't see with other OSS projects |
| 06:44.38 | wltjr | yeah that's good stuff, migrating configs is always fun |
| 06:44.45 | jasonb | Ant used to have oddities like that. They ended up ironing out most of their problems though, shockingly. |
| 06:45.09 | wltjr | I got sick of people bitching when moving from 5.0.x -> 5.5.x because they did not read or pay attention to the changes in logging or etc, and did not update their server.xml |
| 06:45.31 | wltjr | thankfully 5.0.x is gone :) |
| 06:45.39 | wltjr | oldest available is 5.5.23 :) |
| 06:45.54 | jasonb | Really, server.xml hand migration is really harder than we should expect casual users (even sysadmins) to be doing. |
| 06:46.02 | jasonb | (without good docs, that is!) |
| 06:46.20 | wltjr | yes I agree, but once you are familiar with tc, it's not to bad |
| 06:46.28 | jasonb | Me? |
| 06:46.43 | jasonb | It's not a problem for me. :) |
| 06:46.52 | wltjr | jasonb: I agree on docs, but don't think dealing with server.xml by hand is to bad |
| 06:47.08 | wltjr | jasonb: yeah my point, once you are familiar with tc, it's not a big deal |
| 06:47.21 | jasonb | Yeah, it's not bad if you know for sure what you're doing.. but then again nothing really is, in that case. :) |
| 06:47.45 | wltjr | jasonb: for me it's usually just about finding the changes and applying those, but I am sure others get intimidated |
| 06:47.52 | jasonb | "A 2-year old could do it with 18 years experience!" |
| 06:47.56 | wltjr | of course I have my own, WAY slimmed down server xml :) |
| 06:48.36 | wltjr | I do like it, takes me very little to setup a vhost for tc, like 5-6 lines, and linking in root context.xml in Catalina/domain.name.co |
| 06:48.38 | wltjr | m |
| 06:48.56 | jasonb | wltjr: I agree that it's fine for you to have your own subsetted server.xml file.. you can handle that. But, I strongly believe that distros should not subset server.xml and give that out to all users. |
| 06:49.16 | wltjr | oh I don't do that, we ship just as provided by upstream |
| 06:49.22 | jasonb | Good. |
| 06:49.23 | wltjr | unless upstream does not work, then I have to deal |
| 06:49.28 | jasonb | Yeah. |
| 06:49.36 | wltjr | I mean on my servers, dev and production |
| 06:49.46 | wltjr | I don't force my views and habits on gentoo peeps |
| 06:49.54 | jasonb | I tend to hack and slash server.xml as I please, but only because I know very well that I can fix whatever damage I do. :) |
| 06:50.25 | wltjr | in fact lately I have been getting annoyed with one off requests or issues, and am taking more of a hardline if it's not effecting more than one, or benefits more than one, not a high priority if ever considered :) |
| 06:50.50 | wltjr | only thing I can say about tc, is it's not friendly to breakage |
| 06:51.11 | wltjr | like put a broken symlink in one of it's lib dirs, it will fail to start with a VERY vague error |
| 06:51.27 | jasonb | Check out this cool gentoo page: http://viewcvs.gentoo.org/viewcvs.py/gentoo-x86_64/www-servers/tomcat/?hideattic=0 |
| 06:51.28 | jasonb | :) |
| 06:51.34 | wltjr | used to run into that all the time, when I manually installed and linked stuff,, thankfully now it's automated via portage and emerge |
| 06:51.54 | jasonb | I changed it from x86 to x86_64.. I guess you don't support 64-bit x86 processors.. :) |
| 06:52.07 | wltjr | yes, but that's just a cvs root |
| 06:52.15 | wltjr | we support tomcat on freebsd :) |
| 06:52.27 | wltjr | just how the cvs root was created and named |
| 06:52.35 | wltjr | there is another one next to it named gentoo that houses all docs |
| 06:52.42 | wltjr | http://viewcvs.gentoo.org/viewcvs.py/ |
| 06:53.17 | wltjr | you can see the archs here http://packages.gentoo.org/search/?sstring=tomcat |
| 06:53.51 | wltjr | in the ebuilds its controlled by KEYWORD variable, tc 6.0.x is support on ppc now, 5.5.x had issues due to using com.sun.ssl.* stuff not present in non sun-jdks |
| 06:54.14 | wltjr | not sure why ppc64 has not keyworded it yet, other archs can't because they lack java, for now :) |
| 06:54.34 | wltjr | now that openjdk is well open, it will likely be ported to other archs, and will have much wider arch support for java, and java apps |
| 06:54.35 | jasonb | That's such a pretty page. |
| 06:55.29 | wltjr | yeah, and the version pages :) |
| 06:55.29 | wltjr | http://packages.gentoo.org/ebuilds/?tomcat-6.0.13-r3 |
| 06:55.30 | wltjr | http://packages.gentoo.org/images/www-servers/tomcat.jpg |
| 06:55.30 | jasonb | Where can I buy a box of Tomcat? :) |
| 06:55.39 | wltjr | maybe in the amazon or africa? |
| 06:55.57 | wltjr | pretty funny branding though huh |
| 06:56.18 | jasonb | The Tomcat box should really show the Tomcat on the front of the box, instead of some lame little computer. |
| 06:56.30 | wltjr | it's like that for all |
| 06:56.35 | wltjr | http://packages.gentoo.org/ebuilds/?tomcat-native-1.1.10 |
| 06:56.39 | jasonb | No, no, Tomcat is special!! |
| 06:56.43 | jasonb | :) |
| 06:56.52 | wltjr | anything java tends to be hated on |
| 06:56.59 | jasonb | ooh, a red one! |
| 06:57.34 | jasonb | And, near the bottom of the box, it should say something like "Free Catnip Inside!" |
| 06:58.20 | wltjr | if it were only package specific, but it's not, any package in portage will look like that |
| 06:58.44 | wltjr | I wonder if this violates a license http://packages.gentoo.org/ebuilds/?sun-jdk-1.6.0-r2 |
| 06:58.46 | wltjr | :) |
| 07:01.00 | wltjr | when it was around http://web.archive.org/web/20060513022941/http://www.funroll-loops.org/ |
| 07:01.08 | wltjr | used to be pretty funny pun on gentoo :) |
| 07:01.16 | jasonb | :) |
| 07:01.54 | wltjr | we do hate on ricers though |
| 07:02.15 | wltjr | building from source is one thing, trying to rice out apps with insane c/c++ flags is another story |
| 07:02.39 | jasonb | :) |
| 07:02.45 | wltjr | and with that, i am off to bed, 3am on the east coast, rice rice baby :) |
| 07:02.59 | jasonb | Okay, have a good night. TTYL.. |
| 07:03.07 | wltjr | yep l8r all |
| 07:41.20 | *** join/#tomcat codeshepherd (n=codeshep@61.8.208.15) |
| 08:00.34 | *** join/#tomcat prgrmr_ (n=prgrmr@bzq-82-81-87-68.red.bezeqint.net) |
| 08:06.45 | *** join/#tomcat vegard_ (n=vegard@MO061092.fh-weingarten.de) |
| 08:07.39 | vegard_ | does anybody here know how to run the sysdeo-plugin as non-root? |
| 08:08.56 | *** join/#tomcat vegard_ (n=vegard@MO061092.fh-weingarten.de) |
| 08:11.27 | vegard_ | (Tomcat 5.5.23; Java 1.5.0; Ubuntu Feisty) I want to use the sysdeo-plugin in eclipse to control my tomcat-server as non-root. Does anybody know how I can do this? |
| 08:25.46 | *** join/#tomcat durre (n=rocknrol@c-12a8e155.142-3-64736c11.cust.bredbandsbolaget.se) |
| 08:28.25 | *** part/#tomcat durken (n=rocknrol@c-12a8e155.142-3-64736c11.cust.bredbandsbolaget.se) |
| 08:38.03 | *** join/#tomcat gregor_k (n=a@p54a18828.dip0.t-ipconnect.de) |
| 09:02.37 | *** join/#tomcat Tourix (n=Tourix@LPuteaux-151-42-21-52.w193-252.abo.wanadoo.fr) |
| 09:02.51 | Tourix | hi |
| 09:11.23 | *** join/#tomcat [algo] (n=ilia@algo.dialup.corbina.ru) |
| 09:17.43 | *** join/#tomcat janning (n=janning@pd95b4e37.dip0.t-ipconnect.de) |
| 09:25.17 | janning | TOMCAT: 6.0.13 OS: linux-2.6.18-4-686 JAVA: java-1.5.0-sun-1.5.0.10 |
| 09:25.23 | janning | i tried to run with APR, followed http://tomcat.apache.org/tomcat-6.0-doc/apr.html |
| 09:25.41 | janning | after compiling lib starting tomcat with CATALINA_OPTS="-Djava.awt.headless=true -Xmx128M -server -Djava.library.path=/usr/local/apr/lib" |
| 09:25.46 | janning | In my logs tomcat is starting up and shows |
| 09:25.50 | janning | INFO: APR capabilities: IPv6 [true], sendfile [true], accept filters [false], random [true]. |
| 09:25.55 | janning | as last line. No connector starts. When i try "telnet localhost 8080" => Connection refused |
| 09:26.21 | janning | when i drop -Djava.library.path=/usr/local/apr/lib from OPTS it starts up withou APR |
| 09:26.57 | janning | Can someone help me to get APR running? |
| 09:28.49 | *** join/#tomcat Nicke (n=niclasa@ua-83-227-140-135.cust.bredbandsbolaget.se) |
| 09:30.45 | *** join/#tomcat stork (i=stork@storkey.plus.com) |
| 10:01.50 | *** join/#tomcat Boggie (i=Crusader@86.125.238.31) |
| 10:03.44 | Boggie | folks i had to ask. I've deployed some simple web pages to a web application on my tomcat server. all i get about them is 404 messages, unavailable resources. I deployed the files in the tomcat_home/webapps/webapphomedirgoeshere/. I am running Tomcat 6.0.13, with Java 1.6.0 and Windows 2000 pro |
| 10:04.12 | Boggie | any idea why I would get the 404 messages? |
| 10:10.19 | lyken | and how are you trying to access it ? |
| 10:12.32 | Boggie | well with my web browser: http://localhost:<portnumber>/webappname/jsppagename.jsp goesh ere |
| 10:12.48 | Boggie | http://eyeinthesky.homeunix.com:8000/wodgenerator/index.jsp |
| 10:15.49 | Boggie | I read the article at coreservlets.com about tomcat configuration and well I did everything that was said there. |
| 10:17.04 | Boggie | http://eyeinthesky.homeunix.com:8000/ |
| 10:17.11 | Boggie | seems the whole server is down even as it's running |
| 10:20.31 | Boggie | now i'm totally clueless to waht is going on. |
| 10:42.42 | *** join/#tomcat Nevermind90 (n=bub@host246-199-dynamic.7-87-r.retail.telecomitalia.it) |
| 10:42.46 | Nevermind90 | ciao |
| 10:42.50 | *** part/#tomcat Nevermind90 (n=bub@host246-199-dynamic.7-87-r.retail.telecomitalia.it) |
| 10:50.37 | jwisher_ | how is an else statement done withe <c:if test="attr">html</c:if>? |
| 11:12.00 | *** join/#tomcat tim|imac (n=tim@kennelvanstoopy.xs4all.nl) |
| 11:12.36 | tim|imac | hi all, where can i find the reference documentation for tomcat5.5 configuration directives? |
| 11:12.48 | tim|imac | I've tried http://tomcat.apache.org/tomcat-5.5-doc/config/index.html, but it seems incomplete or at least very poorly organised |
| 11:13.02 | tim|imac | I'm looking for directives that are allowed for <Deployer/> |
| 11:26.31 | *** join/#tomcat Nicke (n=niclasa@c83-251-17-30.bredband.comhem.se) |
| 11:53.10 | *** join/#tomcat [algo] (n=ilia@h86-62-117-98.ln.rinet.ru) |
| 12:46.34 | *** join/#tomcat the_sniff (n=sniff@CBL217-132-148-183.bb.netvision.net.il) |
| 12:47.20 | the_sniff | im using tomcat 4, and it runs but doesn't listen on ports |
| 12:48.34 | *** part/#tomcat tim|imac (n=tim@kennelvanstoopy.xs4all.nl) |
| 13:10.59 | the_sniff | anyone alive? |
| 13:11.46 | the_sniff | my tomcat is 4, java1.3.1 debian 4.0 |
| 13:17.50 | *** join/#tomcat wsmoak (n=wsmoak@ip68-110-100-131.ph.ph.cox.net) |
| 13:18.38 | *** part/#tomcat janning (n=janning@pd95b4e37.dip0.t-ipconnect.de) |
| 13:36.22 | *** join/#tomcat Nicke (n=niclasa@p35.kthopen.kth.se) |
| 13:45.57 | *** join/#tomcat randrew (n=raj@dolmen.cc.columbia.edu) |
| 13:55.03 | *** join/#tomcat CrummyGummy (n=CrummyGu@dsl-242-23-15.telkomadsl.co.za) |
| 13:55.03 | *** join/#tomcat awelynant (n=cskinfil@208.65.172.147) |
| 13:57.26 | *** join/#tomcat Nicke (n=niclasa@p45.kthopen.kth.se) |
| 14:24.00 | *** join/#tomcat melter (n=Melter@dyn007.morrill-116.ndsu.NoDak.edu) |
| 15:03.49 | *** join/#tomcat DSpair (n=dphillip@12.44.222.61) |
| 15:04.22 | DSpair | I have a Tomcat server running 5.5, and I have created a class called ProcessExcel which takes a multipart/form-data post from the browser. |
| 15:04.36 | DSpair | When I submit the form from the JSP, I get a 405 "HTTP method POST is not supported by this URL" |
| 15:05.03 | DSpair | I have checked my security constraints and I do have a "doPost" method, but still no joy... Any suggestions? |
| 15:15.18 | *** join/#tomcat pappy-_ (n=alex@dslb-084-059-017-214.pools.arcor-ip.net) |
| 15:23.01 | *** join/#tomcat pmcmahan (i=pmcmahan@nat/ibm/x-7c3bd91cce0ddea8) |
| 15:26.12 | randrew | DSpair: check that the URL maps to the servlet you expect. did you inadvertently create a service() method? |
| 15:42.48 | *** join/#tomcat fix_ (n=fix@cable-87-244-191-178.upc.chello.be) |
| 15:46.52 | *** join/#tomcat bdc34 (n=bdc34@caruso.mannlib.cornell.edu) |
| 15:52.13 | *** join/#tomcat karstensrage_alt (n=karstens@c-69-181-220-5.hsd1.ca.comcast.net) |
| 15:56.34 | *** join/#tomcat roel-_ (i=rdvdijk@rustradio.xs4all.nl) |
| 15:59.23 | *** join/#tomcat DSpair (n=dphillip@12.44.222.61) [NETSPLIT VICTIM] |
| 16:05.01 | *** join/#tomcat CrummyGummy (n=CrummyGu@dsl-242-23-15.telkomadsl.co.za) [NETSPLIT VICTIM] |
| 16:05.01 | *** join/#tomcat pmcmahan (i=pmcmahan@nat/ibm/x-0133344199e54248) [NETSPLIT VICTIM] |
| 16:05.29 | *** join/#tomcat awelynant (n=cskinfil@208.65.172.147) [NETSPLIT VICTIM] |
| 16:10.39 | *** join/#tomcat rambobo1 (n=king@host167-72-dynamic.4-79-r.retail.telecomitalia.it) |
| 16:16.55 | *** join/#tomcat ramboboa (n=king@host167-72-dynamic.4-79-r.retail.telecomitalia.it) |
| 16:17.14 | ramboboa | hello,i have a trouble with tomcato 6.0.13 |
| 16:17.36 | ramboboa | i install it but i can-t log into admin and manager pages... |
| 16:20.47 | *** part/#tomcat ramboboa (n=king@host167-72-dynamic.4-79-r.retail.telecomitalia.it) |
| 16:31.49 | *** join/#tomcat Arathorn (n=Arathorn@puma.mxtelecom.com) |
| 16:34.47 | Arathorn | hi all - on tomcat 5.5.20, java 1.4.2_09, and linux (libc 2.3.2 on i386): I have a production tomcat which is using stale JARs for a particular webapp, despite the webapp's WAR having been redeployed several times with newer JARs within it. |
| 16:35.07 | Arathorn | on reading /proc/<pid>/maps for the tomcat parent process |
| 16:35.26 | Arathorn | I see that the mmaps for the JARs in question are listed as 'deleted' |
| 16:35.37 | Arathorn | and indeed refer to an inode for a file which is no longer on the system |
| 16:36.22 | Arathorn | e.g. 7d358000-7d3af000 r--s 00000000 08:01 17711409 /home/appstomcat/webapps/foo/WEB-INF/lib/mail-1.3.3.jar (deleted) |
| 16:36.51 | Arathorn | (whereas the inode number for the current mail-1.3.3.jar at that path is 16941705) |
| 16:37.07 | Arathorn | what would cause tomcat to fail to pick up new versions of a handful of JARs? |
| 16:37.39 | DSpair | Arathorn: Have you checked to see if the old version of the jars is in ${tomcat_home}/common/lib or ${tomcat_home}/server/lib? |
| 16:37.59 | DSpair | They get precedence on class loading. |
| 16:38.14 | Arathorn | yup, none of the 5 jars are present elsewhere on the system |
| 16:38.23 | Arathorn | especially not in common/lib or server/lib |
| 16:38.39 | Arathorn | (they're all fairly specific - e.g. wicket-1.3.0-incubating-SNAPSHOT.jar as well as mail-1.3.3.jar) |
| 16:38.47 | DSpair | Do you have the original classes for the jars in your WEB-INF/classes folder? |
| 16:39.18 | DSpair | The unpackaged classes would also get preference. |
| 16:39.49 | Arathorn | no, i don't |
| 16:40.02 | DSpair | Hrrrmmmmm..... Have you tried restarting tomcat? |
| 16:40.09 | Arathorn | i haven't, as it's a production box |
| 16:40.25 | Arathorn | and it's not very convenient to do so right now, and i'd prefer to actually understand why tomcat's got wedged :) |
| 16:40.30 | DSpair | Aside from that I have no suggestions... Have you tried it on another tomcat installation? |
| 16:41.07 | Arathorn | 3 tomcats in the same cluster are all identically wedged |
| 16:41.15 | DSpair | Arathorn: Were those jars ever in the server/lib or common/lib? To release those jars from the classloader, you have to restart the application server. |
| 16:41.38 | Arathorn | they will never have been in {server,common}/lib - only ever deployed as part of a WAR |
| 16:41.43 | Arathorn | although i see your point |
| 16:41.59 | DSpair | Arathorn: I'm out of ideas bud... |
| 16:42.24 | Arathorn | 'tis frustrating. any idea how heavy it is to run a tomcat with -Xdebug for remote debugging? |
| 16:42.24 | DSpair | I would install tomcat on my workstation and try it there. |
| 16:42.45 | DSpair | Arathorn: Not TOO bad, but I wouldn't recommend it for a high volumne production site. |
| 16:42.55 | DSpair | s/volumne/volume/ |
| 16:43.16 | Arathorn | that prolly counts this out, then |
| 16:43.21 | DSpair | Stupid bot.. He knows what I meant. |
| 16:44.13 | Arathorn | the thing is - almost all the new JARs in the WAR get picked up, so deploying it on another container is going to be the same as restarting the existing one |
| 16:44.22 | Arathorn | (as it won't have any stale versions to fall back to) |
| 16:44.49 | DSpair | Arathorn: Agreed, but that will at least give you an indication as to if the classes will load at all. |
| 16:44.52 | Arathorn | i just wondered if anyone had experienced stale JARs getting wedged in the classloader ho hum. |
| 16:44.54 | Arathorn | true. |
| 16:45.25 | DSpair | Arathorn: Another option would be to use a manual class loader... |
| 16:45.53 | Arathorn | within the webapp itself? |
| 16:45.56 | DSpair | Arathorn: you could write code in your servlet context to specifically load a jar file from specifc location. |
| 16:46.48 | DSpair | Arathorn: Yes, I have seen it done in MANY production environments. Instead of reloading the WAR file, send an RMI request to the Servlet Context to run a class loader thread and refresh your classes to the newer files. |
| 16:47.13 | Arathorn | mm, interesting |
| 16:47.19 | Arathorn | nice idea |
| 16:47.26 | DSpair | Arathorn: For those times when you absolutely cannot take your app off-line. |
| 16:47.48 | DSpair | I can't take credit for it. I have seen other people do it before./ |
| 16:48.57 | DSpair | Arathorn: You say you have an application server cluster? Why can't you just restart one node at a time? |
| 16:49.12 | Arathorn | well, i can ;) |
| 16:49.22 | Arathorn | but i still want to try to understand why it wedged whilst it's still wedged |
| 16:49.36 | DSpair | Aha.... Nothing in the logs? |
| 16:49.47 | Arathorn | nothing to do with those jars specifically |
| 16:50.04 | DSpair | You checked catalina.out? |
| 16:50.41 | Arathorn | yup |
| 16:50.44 | pfn | huh? tomcat wedged? say what? |
| 16:50.45 | Arathorn | plus the various log4j destinations |
| 16:50.56 | pfn | and hot deployment in a production server? say what? |
| 16:51.02 | Arathorn | there's one rather painful failure to load the context for this particular webapp a few days ago |
| 16:52.01 | Arathorn | which manifested itself as a classic |
| 16:52.02 | Arathorn | SEVERE: Error listenerStart / 29-May-2007 11:33:26 org.apache.catalina.core.StandardContext start / SEVERE: Context [/foo] startup failed due to previous errors |
| 16:52.05 | Arathorn | in catalina.out |
| 16:52.16 | Arathorn | (with an associated exception stacktrace elsewhere) |
| 16:52.28 | Arathorn | but i can't justify to myself why that'd cause those jars to wedge. |
| 16:52.30 | pfn | then the actual error is in localhost.log |
| 16:52.35 | pfn | what do you mean "wedge" |
| 16:53.25 | Arathorn | the actual error is actually wherever the log4j configured it to go (a remote appender, as it happens) |
| 16:53.41 | Arathorn | by wedge, i mean that the tomcat is now using stale versions of the jars for the webapp in question |
| 16:53.50 | pfn | how are you certain of that? |
| 16:54.08 | pfn | and why are you updating jars in WEB-INF/lib anyway? (not that it should cause a problem) |
| 16:54.20 | Arathorn | because a) the behaviour is from the old jars, and b) looking in /proc/<pid>/maps shows it to be still memory mapping deleted versions of the jars in question |
| 16:54.31 | pfn | then reload the app again |
| 16:54.33 | Arathorn | i'm not updating them manually, i'm just deploying new versions of a WAR |
| 16:54.39 | pfn | and again, hot deployment is never recommended for a production server |
| 16:54.41 | Arathorn | it's been redeployed several times. |
| 16:54.48 | Arathorn | how am I hot deploying? |
| 16:54.48 | pfn | regardless of app server/servlet container |
| 16:54.56 | pfn | hot deploying = hot redeploying |
| 16:55.01 | pfn | = deploying new version of a jar |
| 16:55.14 | Arathorn | but i'm not deploying a new version of a jar - i'm redeploying the whole webapp WAR |
| 16:55.25 | pfn | a war is a jar |
| 16:55.38 | Arathorn | surely tomcat's intended to let you replace WARs whilst the container's running, without restarting the whole server? |
| 16:55.41 | pfn | anyway, my meaning in the above statement is war |
| 16:55.51 | pfn | arathorn it's never recommended in production mode, not on any server |
| 16:55.55 | Arathorn | interesting |
| 16:55.56 | pfn | not on websphere, not on weblogic |
| 16:55.57 | Arathorn | didn't realise that |
| 16:56.19 | pfn | but even so, you should not experience a problem as you're describing |
| 16:57.47 | Arathorn | any suggestions on how I might debug it at all, as the tomcat isn't running in debugging mode |
| 16:57.49 | pfn | anyway, if the old jar is still around, that's not surprising if you have a reference to a classloader somewhere |
| 16:58.09 | pfn | although new classes should never reference the old jar |
| 16:58.19 | pfn | as all new classloaders are created when you drop the new war in |
| 16:58.24 | Arathorn | hm, the only classloader which could have a reference to it would be the container's classloader, presumably? |
| 16:58.46 | DSpair | Sorry about that... I |
| 16:58.57 | pfn | cat /proc/PID/maps | grep jarnameinquestion and pastebin it |
| 16:59.04 | pfn | also ls -i jarnameinquestion |
| 16:59.20 | DSpair | 'I'm rendering Digital Video on my other screen and it locked up my XWindows display (I shouldn't renice to -10).... :) |
| 17:00.23 | Arathorn | 7ed00000-7ed4e000 r--s 00000000 08:01 8470779 /home/appstomcat/webapps/foo/WEB-INF/lib/wicket-extensions-1.3.0-incubating-SNAPSHOT.jar (deleted) |
| 17:00.24 | Arathorn | 16941714 /home/appstomcat/webapps/foo/WEB-INF/lib/wicket-extensions-1.3.0-incubating-SNAPSHOT.jar |
| 17:00.57 | Arathorn | 8470779 != 16941714 :( |
| 17:04.00 | pfn | that's the only entry in the list? |
| 17:04.07 | pfn | reload the app from the manager |
| 17:06.57 | Arathorn | that was the only entry (the only output of cat /proc/1682/maps | grep wicket-extensions-1.3.0-incubating-SNAPSHOT.jar |
| 17:06.58 | Arathorn | <PROTECTED> |
| 17:07.47 | *** join/#tomcat [algo] (n=ilia@h86-62-117-98.ln.rinet.ru) |
| 17:10.14 | Arathorn | except the manager's proxy is broken - brb |
| 17:18.50 | *** join/#tomcat rambobo1 (n=king@d83-176-10-78.cust.tele2.it) |
| 17:25.50 | rambobo1 | hello,i have this problem with tc...i am on the browser,but i cant find manager and admin page...why |
| 17:29.18 | evosh | pfn, Reading your comments, I should ask, what are the other alternatives to redeploy a library on a production environment? |
| 17:29.36 | pfn | evosh restart the server |
| 17:29.40 | pfn | if it's a cluster, that's not a problem |
| 17:29.48 | evosh | Having multiple Tomcat instances as a cluster, shutting one down, updating, reloading, moving on to the next instance? |
| 17:29.55 | pfn | just update your lb configuration not to pass to the server and once the session count is down to 0 |
| 17:29.55 | evosh | I see. |
| 17:29.58 | pfn | restart the server |
| 17:30.04 | pfn | if you need to get the change out |
| 17:30.12 | pfn | at a certain time |
| 17:30.16 | pfn | then you create a maintenance window |
| 17:30.19 | pfn | and shutdown briefly |
| 17:30.29 | Arathorn | so the only usage of a container then becomes clustering? |
| 17:30.30 | evosh | Same goes for a single instance. |
| 17:30.45 | pfn | Arathorn, huh? |
| 17:30.52 | pfn | the only usage of a container? |
| 17:31.39 | Arathorn | well, other than patching the HTTP requests/responses into the servlets API |
| 17:32.04 | Arathorn | i just always thought of hot-deploying as a fairly fundamental feature of a container :) |
| 17:32.44 | evosh | Ok, on a single Tomcat instance running server, is it possible to introduce a new instance, migrate sessions from the first instance to the second, shut down the first, do the maintenance, then revive it? |
| 17:33.09 | pfn | arathorn no, the http request/response is the fundamental aspect of a container |
| 17:33.10 | evosh | Arathorn, Same here. |
| 17:33.20 | pfn | hot deployment is only a development feature |
| 17:33.46 | Arathorn | pfn: reloading it through the manager app didn't fix it |
| 17:33.46 | pfn | evosh that doesn't really work... you could do something like that, though, run another instance of tomcat on another port/IP |
| 17:34.01 | pfn | arathorn I don't know what's wrong for you then |
| 17:34.02 | Arathorn | the deleted jars are still taking precedence over the ones in the actual WAR |
| 17:34.06 | Arathorn | nor I :( |
| 17:34.17 | pfn | arathorn and PID/maps doesn't list it? |
| 17:34.18 | *** join/#tomcat golan (n=golan@momona3.momona.org) |
| 17:34.19 | pfn | just restart... |
| 17:34.20 | evosh | Something is fishy here then. |
| 17:34.35 | pfn | arathorn of course, last question, the unpacked jar files are readable to tomcat? |
| 17:34.37 | Arathorn | PID/maps only lists the deleted one. |
| 17:34.57 | evosh | Hah, don't tell me it's a permission issue! |
| 17:35.14 | Arathorn | no, perms are all fine |
| 17:35.27 | Arathorn | 0644, owned by the tomcat user |
| 17:35.35 | Arathorn | likewise the parent dirs. |
| 17:35.52 | pfn | in development, I've been redeploying my app over and over, no jars get wedged... |
| 17:36.19 | pfn | b51b6000-b51d9000 r-xs 0015e000 08:03 8636323 /usr/local/apache-tomcat-6.0.13/webapps/wicket-examples/WEB-INF/lib/wicket-1.3.0-incubating-SNAPSHOT.jar |
| 17:36.22 | pfn | b5301000-b5324000 r-xs 0015e000 08:03 8637502 /usr/local/apache-tomcat-6.0.13/webapps/reseller-crm/WEB-INF/lib/wicket-1.3.0-incubating-SNAPSHOT.jar (deleted) |
| 17:36.51 | evosh | What's a wedged jar? Not redeployed? Stale lib still being reference? |
| 17:37.23 | pfn | hmm, oops, looks like this is the wrong jar |
| 17:37.23 | pfn | stupid wicket examples |
| 17:37.27 | evosh | Hah. |
| 17:37.47 | Arathorn | it happens with that JAR too |
| 17:37.59 | pfn | hmm, interesting, looks like tomcat doesn't reload jars in WEB-INF/lib |
| 17:38.38 | Arathorn | during a redeploy just then, i noticed that it had both versions mapped, actually |
| 17:38.44 | Arathorn | both the deleted one and the new one |
| 17:39.15 | Arathorn | 7df90000-7e108000 r--s 00000000 08:01 12468850 /home/appstomcat/webapps/foo/WEB-INF/lib/wicket-1.3.0-incubating-SNAPSHOT.jar (deleted) |
| 17:39.15 | Arathorn | 81df7000-81f78000 r--s 00000000 08:01 16941961 /home/appstomcat/webapps/foo/WEB-INF/lib/wicket-1.3.0-incubating-SNAPSHOT.jar |
| 17:39.16 | Arathorn | like so |
| 17:39.34 | pfn | http://radio.weblogs.com/0135826/2004/05/17.html#a30 |
| 17:39.36 | pfn | anyway, this is old |
| 17:39.41 | pfn | but a general commentary on hot deployment |
| 17:39.52 | Arathorn | so is it possible that it's mapping it for long enough to think it doesn't need to replace it with the newer version? |
| 17:43.00 | Arathorn | hmm, is there any chance that "INFO: validateJarFile(/home/appstomcat/webapps/foo/WEB-INF/lib/jsp-api-2.0.jar) - jar not loaded. See Servlet Spec 2.3, section 9.7.2. Offending class: javax/servlet/jsp/JspPage.class" might cause other jars to not reload? |
| 17:43.14 | Arathorn | evosh: by wedged, i mean stale lib still being referenced despite a full redeploy |
| 17:43.28 | pfn | arathorn don't know, maybe |
| 17:44.08 | pfn | why do you have jsp-api in your WEB-INF/lib |
| 17:44.08 | pfn | tomcat already provides it |
| 17:44.08 | pfn | doing stuff like *that* is a surefire way to get problems |
| 17:45.53 | Arathorn | possibly so that it also deploys on jetty? |
| 17:47.22 | Arathorn | apparently it's so that taglibs can build against it |
| 17:47.34 | pfn | on jetty, you should just include jsp-api in jetty's classpath |
| 17:47.41 | pfn | you shouldn't do that on any container that has jsp support |
| 17:47.54 | pfn | you'll run into problems on jboss, glassfish, websphere, weblogic, etc. by including an incompatible jsp-api like so |
| 17:48.13 | Arathorn | interestingly, whilst the webapp deploys, for about 30 seconds the correct new JARs are mapped (as well as the deleted ones) |
| 17:48.18 | Arathorn | but then it unmaps |
| 17:48.31 | pfn | so remove the jsp-api.jar and any other conflicting file and see what happens |
| 17:50.01 | pfn | your ant script :p |
| 17:51.59 | Arathorn | indeed - i didn't write it though ;) |
| 17:52.18 | evosh | I see. Interesting, I dump a new mysql jdbc driver in WEB-INF/lib and doesn't auto-redeploy it. |
| 17:53.22 | evosh | As a matter of fact, it doesn't even load it (!!) |
| 17:53.25 | Arathorn | does it load it when you redeploy the app using it? |
| 17:53.37 | Arathorn | and when you say 'new', do you mean that it has the same filename but newer content? |
| 17:54.11 | evosh | No and yes. |
| 17:55.00 | evosh | Let me move it to common/lib and restart the server. |
| 17:56.23 | pfn | evosh you can't put jdbc drivers into WEB-INF/lib |
| 17:56.28 | pfn | not if you want the container to manage connections |
| 17:58.36 | evosh | Well, I just wanted to see if I can redeploy it from WEB-INF/lib. |
| 17:59.54 | evosh | Ok, stupid of me, I had renamed it myslq (mysql). Pfft. |
| 17:59.54 | evosh | Took all the flavor out of it. |
| 18:05.25 | *** part/#tomcat rambobo1 (n=king@d83-176-10-78.cust.tele2.it) |
| 18:08.28 | *** join/#tomcat svanga (n=chatzill@192.18.43.225) |
| 18:19.38 | Arathorn | meh, i wish i could debug this |
| 18:32.18 | *** join/#tomcat [algo] (n=ilia@h86-62-117-98.ln.rinet.ru) |
| 18:42.57 | *** join/#tomcat bdudney (n=bdudney@207.182.253.37) |
| 19:23.09 | *** join/#tomcat columbo (n=Miranda@pool-71-104-45-67.lsanca.dsl-w.verizon.net) |
| 19:53.55 | *** join/#tomcat svanga_ (n=on@adsl-69-226-230-248.dsl.pltn13.pacbell.net) |
| 19:58.43 | *** join/#tomcat prgrmr (n=prgrmr@bzq-82-81-87-68.red.bezeqint.net) |
| 21:01.14 | *** join/#tomcat wsmoak_ (n=wsmoak@ip68-110-100-131.ph.ph.cox.net) |
| 21:09.12 | *** join/#tomcat LongBeach (n=mike@AFontenayssB-152-1-58-100.w82-121.abo.wanadoo.fr) |
| 21:44.11 | *** join/#tomcat vinse (n=vinse@208.253.223.146) |
| 22:38.21 | minerale | when editing a jsp page I will switch between my browser and eclipse. When the jsp breaks down the refreshes on the browser seem erratic, one in four refereshes returns an error pages while the others return a live page |
| 22:38.35 | minerale | I can't figure out if it's tomcat that's doing this or if it is the browser showing me some cache |
| 23:36.13 | jwisher_ | what happens if someone posts if you have a form where action="/muchlawl/lol.jsp?lol=truetrue" method="post" ? what does request.getParameter do if you had a form element also called lol? |
| 23:54.41 | pfn | you get screwed |
| 23:55.02 | pfn | you can try getParameterValues("lol") which returns a String[] |
| 23:55.04 | pfn | or something like that |
