That's a long live problem which exists in all
Quercushttp://caucho.com/resin/doc/quercus.xtp
version.
I believe that's because of the poor unicode implementation of mysql (or all
db specific) functions.
I raised the question to Scott one years before, yet not solved now.
- Wesley
2009/10/6
Thanks Scott.
I'll test the JPA JSR299/330 stuff.
2009/10/14 Scott Ferguson f...@caucho.com
The 4.0.2 snapshot is now available. Bugs can be reported at
bugs.caucho.com. The snapshot isn't clean yet, so a bug report might
duplicate one of our failing regression tests, but it's better to
Seems 4.0.2 snapshot passed most of my web test.
Passed:
Thirdparty Hibernate JPA integration (* see issues below)
@Inject
@Qualifier
@Observes (with @Qualifier bindings)
@ApplicationScoped
@RequestScoped
Failed:
@SessionScoped
not tested:
@InterceptorBinding
I found two issues so far:
1. I
I think I could only use @SessionScoped annotation on SINGLE beans,
though not tested yet.
sorry, I mean SIMPLE beans not SINGLE.
-Wesley
2009/10/14 Wesley Wu wumen...@gmail.com
Seems 4.0.2 snapshot passed most of my web test.
Passed:
Thirdparty Hibernate JPA integration (* see issues below
Hi Scott,
I got two new problems.
I have a scheduled task need to inject a @PersistenceUnit, as the config
below:
resin:ScheduledTask xmlns:tasks=urn:java:com.timehr.tasks
resin:cron0 2 */resin:cron
resin:task
tasks:PositionRefreshTask/
/resin:task
/resin:ScheduledTask
I turned log level to
You may use http://tuckey.org/urlrewrite/ UrlRewriteFilter.
I wrote a similar filter doing the same thing which loads rewrite config
from database.
2009/10/30 Riccardo Cohen r...@architectedulogiciel.fr
Hello
I didn't have yet the opportunity to work with search engine friendly
urls with
syntax. So
if my controller parses the url /servlet?command=showuserid=1354, why
couldn't it parse the url /user/name instead directly ? This removes
from the application one level of control and complexity .
Am-I right ?
Wesley Wu wrote:
Not recommended.
I think filter should
http://www.caucho.com/resin/changes/changes.xtp
[show] javascript:show();
/var/www/hosts/www.caucho.com/webapps/resin/changes/changes.xtp:116:
expected
name at ` '
114: liquercus: Drupal and OpenID (#3609, rep by B. Wu)/li
115: liquercus: QuercusParseException - missing semicolon within a
. CDI is actually based on
JBoss Seam and Google Guice.
Cheers,
-Kai
On 27.11.2009, at 12.10, Wesley Wu wumen...@gmail.com wrote:
As long as Scott continue to work for the Resin's future I'll never
switch to other platform.
I loved the WebBeans so much from December 2007, now called
This bug should be marked as block because it prevents resin load-balancer
and third-party web server integration from working properly.
It seems caucho QA team did not pay much attention to Hmux stuff, because I
suffered a lot from it...
-Wesley
2009/11/29 Alex a...@caucho.com
Multi-Part
Without this bug fixed, 4.0.2 can't be used in production environment.
I use two machines with two resin as a load balance cluster, one with
a web-tier and a app-tier the other only a app-tier.
Every file upload block it.
2009/11/30 Wesley Wu wumen...@gmail.com
This bug should be marked
by then, but a snapshot with
fixes will be ok.
Or I'll have to switch back to the 4.0.2 snapshot 091013 (which has
performance/logging issue and JPA issues).
2009/12/1 Scott Ferguson f...@caucho.com:
Scott Ferguson wrote:
Wesley Wu wrote:
Without this bug fixed, 4.0.2 can't be used in production environment
:
Wesley Wu wrote:
Another SEVERE issues with this snapshot.
All POST request was dramatically CHANGED to GET request.
I'm not seeing that behavior. Can you turn on the finer logging and see
what filters/includes/forwards are happening?
-- Scott
-Wesley
2009/12/3 Scott Ferguson f
I think the win32/win64 dlls should be recompiled against the new modification.
2009/12/3 Wesley Wu wumen...@gmail.com:
I wonder why this happened.
I encountered this in all 4.0.2 versions in my developer machine while
not occurred in production machine.
I'll do some investigation next
/3.5.5 GTB6 (.NET CLR 3.5.30729)
127.0.0.1 - - [03/Dec/2009:08:18:21 +0800] GET /css/style05.css
HTTP/1.1 304 - http://localhost:889/register/; Mozilla/5.0
(Windows; U; Windows NT 6.0; zh-CN; rv:1.9.1.5) Gecko/20091102
Firefox/3.5.5 GTB6 (.NET CLR 3.5.30729)
-Wesley
2009/12/3 Wesley Wu wumen
that GET was the subsequent redirect after the POST.
I use redirect after post pattern.
A request first POST to /register and after some processing,
redirect to /register again using a GET request.
-Wesley
2009/12/3 Scott Ferguson f...@caucho.com:
Wesley Wu wrote:
I changed my post url
Hi Scott,
I'm deploying Resin 4.0.2 snapshot to a cluster and found my
@ApplicationScoped bean not distributed across the cluster nodes.
Is there anything to config to make it possible?
I searched Gavin's blog and found some examples regarding @Observes to
resolve this situation.
If I
[09-12-18 01:43:05.221] {hmtp-baa-to-aaa-0}
com.caucho.hemp.broker.HempMemoryQueue consumeQueue
java.lang.IllegalStateException: WebSocket binary must begin with a
0x80 packet at 0x51 (Q)
[09-12-18 01:43:05.221] {hmtp-baa-to-aaa-0} at
this node is using Resin-4.0.s091216
the other is using Resin-4.0.s091202
-Wesley
___
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest
Thanks. I switched back to 091202 snapshot and this issue was resolved.
Would u take some time to look at my other mail about
@ApplicationScoped bean distribution?
-Wesley
2009/12/18 Scott Ferguson f...@caucho.com:
Wesley Wu wrote:
this node is using Resin-4.0.s091216
the other is using
So if I want my beans synchronized across the cluster, I have to use
either JMS or some thirdparty cluster middleware like JBossCache?
-Wesley
___
resin-interest mailing list
resin-interest@caucho.com
, there is an oss free version) = love.
I'm not super clear on what you want, but it sounds like you want the
TIM-MasterWorker (ExecutorService):
http://forge.terracotta.org/releases/projects/tim-messaging/docs/about.html
jon
On Thu, Dec 17, 2009 at 11:51 AM, Wesley Wu wumen...@gmail.com wrote:
So if I
Thanks Jon, I'll definitely give terracotta a try.
As far as I know, EHCache was a opensymphony project one or two years
ago. I noticed that ehcache was acquired by terracotta and became a
key component of terracotta.
That's great!
-Wesley
___
My log output this error:
[09-12-23 04:16:12.154] {hmtp-baa-to-aaa-5}
com.caucho.hemp.broker.HempMemoryQueue consumeQueue
java.lang.NullPointerException
[09-12-23 04:16:12.154] {hmtp-baa-to-aaa-5} at
com.caucho.server.cluster.HmuxQueue.authenticate(HmuxQueue.java:263)
[09-12-23 04:16:12.154]
Hi Jon Scott,
I don't like apache either but resin 4.0.2 cluster web-tier seems
unstable for me.
I've not tested the 4.0.3 cluster.
My Apache config:
ResinHost 192.168.1.4 6801
ResinBackup 192.168.1.5 6801
Which load balancer will be more appropriate for this usage?
Thanks.
-Wesley
Hi Scott,
When a @ApplicationScoped/@Singleton bean is designed to be initialized
after webapp starts (I.E. when first http request was fired) instead of being
initialized during startup stage, in given circumstance, it won't be
injected correctly.
This is my test case:
== StartBean
/29 Scott Ferguson f...@caucho.com:
Wesley Wu wrote:
Hi Scott,
When a @ApplicationScoped/@Singleton bean is designed to be initialized
after webapp starts (I.E. when first http request was fired) instead of being
initialized during startup stage, in given circumstance, it won't be
injected
according to the JSR299 JSR330 final spec
should use javax.inject.Inject instead
2010/2/7 smallufo small...@gmail.com:
I am migrating from Spring to JSR-299 ,
I followed this http://blog.caucho.com/?p=137 ,
but the first frustration is that I cannot find @Current .
I extracted
the docs should be updated
I can't agree more. :)
2010/2/7 smallufo small...@gmail.com:
Oops , thanks for replying so soon.
BTW , the docs should be updated ...
2010/2/7 Wesley Wu wumen...@gmail.com
according to the JSR299 JSR330 final spec
should use javax.inject.Inject instead
2010/2
This could be a lazy init problem I think.
The scheduled tasks would be inited before some of other beans which
the tasks need.
I've met this before.
The workaround:
Inject the Injector into your task, no other webbeans.
When the task starts (run()), create webbeans instances via the injector.
Not a version issue.
An entitymanager should not get transaction by itself. Transaction in
a modern java appserver should be XA or JTA transaciton.
An entitymanager will detect if there is a JTA transaction existing
and will join it if there is an open one.
You need to use an injected
To make set method auto translated into a UPDATE clause, the
entitymanager should be opened after a transaction begins.
@PersistenceUnit(unitName=example)
EntityManagerFactory emf;
EntityManager em;
try {
ut.begin();
EntityManager em= emf.createEntityManager();
CourseBean
@TransactionAttribute.
I never inject a @PersistentContext or never use a container provided
EntityManager.
I use ThreadLocal to maintain every EntityManager instance.
-Wesley
2010/3/30 Stargazer starga...@blueyonder.co.uk
On 30-Mar-2010 09:34, Wesley Wu wrote:
To make set method auto
I will not call UserTransaction.begin() when all db operations are SELECT.
I create a EntityManager from emf to do SELECT stuff with no transaction at
all.
When I detected there're some update/delete operations, I'll do this:
1. close the former created EntityManager if there is one (for
More to mention about MDBs (Message Driven Beans):
1. Always use @TransactionAttribute(TransactionAttributeType.NOT_SUPPORTED)
on public void onMessage(Message message).
because the impl of Resin MDBs has severe reenter synchronization problem.
If I start a transaction before onMessage() being
Hi,
I noticed there is a 4.0.8 release at 6/30, and I tried it out.
Seems fine besides some minor problem described below.
***. MDBs implements MessageListener can't have an @Override on
onMessage method.
I have to remove these annos to make it run.
@Override
Hi Rick
Try to place the tld at WEB-INF/classes/META-INF/your_tld_file.tld.
-Wesley
___
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest
-version1.0/tlib-version
...
/taglib
has not problem in classes/META-INF
2010/7/5 Rick Mann rm...@latencyzero.com:
On Jul 4, 2010, at 21:55:34, Wesley Wu wrote:
Hi Rick
Try to place the tld at WEB-INF/classes/META-INF/your_tld_file.tld.
I don't think that should be necessary. You're supposed
Hi Scott,
The request:
Make caucho apache module to act (maybe should be configurable) as below :
When check a Resin instance whether it is ready to accept connections,
make sure it's not only started normally, but also all CDI startup beans were
already initialized and deployed (after
Thanks for the Priority = normal :)
-Wesley
___
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest
Oh my god Scott, you rule it!
/**
* Returns the current environment container.
*/
public static InjectManager getCurrent(ClassLoader loader)
{
return _localContainer.get(loader);
}
finally the synchronized (_localContainer) was gone!
You have no idea how I suffered from the
Hi Scott,
I found a frustrating problem when porting my framework to accommodate the
Servlet 3.0 file upload spec in Resin 4.0.8/4.0.9.
There did has a (private Object _value) in
com.caucho.server.http.HttpServletRequestImpl.PartImpl, however,
if it's not a file upload but a common text form
Hi Scott,
We're running Resin 4.0.5 for a heavy traffic site (100K+ Pageviews in an
hour) for a couple of months.
In the past several days I've experienced frequent failing responding of
requests after two or three hours a Resin instance was started.
The environment:
Hardware: 16 cores with
Hi Scott,
I've experienced weird connection pool issue in 4.0.9 these days.
After every hour (exactly), the 4.0.9 resin refused to service any
request and occasionally report ConnectionPool [...] pool overflow.
This situation never occurred in prior resin releases and neither my
webapp nor
Thanks. Be sure to notify in the mailing list when the snapshot is out. :)
2010/8/7 Scott Ferguson f...@caucho.com:
Wesley Wu wrote:
Hi Scott,
I've experienced weird connection pool issue in 4.0.9 these days.
After every hour (exactly), the 4.0.9 resin refused to service any
request
Hi Scott,
I've applied the 4.0.10 snapshot and the Alarm issue went away. Thanks.
But Resin consumed all memory after certain hours and resulted in a
halt or restart.
I did a jrockit flight recording and found there was a slow heap
increase during various GCs.
The recording file shows the
Hi Jan,
Did your resin server have a busy traffic? Did u observed any memory
leak or heap overflow?
-Wesley
___
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest
Hi Scott,
The memory leak circumstance was not spotted in Resin 4.0.5 and early
4.0.x versions.
One of the inject path is:
* A filter CmsPageFilter dispatch a request to a Jsp page through
javax.servlet.RequestDispatcher.forward(request, response)
* Jsp page called a BeanMethod custom jsp tag
*
My BeanManager wrapper code
@Singleton
@Startup
public class MDIObjectFactory {
private static BeanManager beanManager;
private static BeanManager getBeanManager() {
if (beanManager == null) {
try {
Hi Scott,
Thanks for the checking.
Today I double checked the CDI spec and got some new knowledge.
Nearly all my beans were not annotated with any scope, so that they
should be @Dependent.
Some of the beans were created (not injected) in a servlet filter via
a static ObjectFactory.
I think the
modifying the HashMap. It's an easy fix.
-- Scott
Here is a more detailed explanation:
http://forums.sun.com/thread.jspa?threadID=5335803
Aaron
On 8/19/2010 4:20 AM, Wesley Wu wrote:
Often happened at 30 seconds after appserver start.
[10-08-19 17:11:44.378] {server://*:6801-487
Hi Scott,
I believe I've fixed this issue by adding @ApplicationScoped and
@RequestedScoped annotation to my beans. :)
Thanks for your efforts.
-Wesley
2010/8/18 Wesley Wu wumen...@gmail.com:
Hi Scott,
Thanks for the checking.
Today I double checked the CDI spec and got some new knowledge
2010/8/20 Scott Ferguson f...@caucho.com:
I still think there's a chance for a Resin leak; I just don't understand
where it's coming from.
I think so. The same code worked fine (no leak) in Resin 4.0.5.
It shouldn't matter what context you're injecting from, as long as
you're creating a new
I'll tried to do so.
2010/8/20 Scott Ferguson f...@caucho.com:
Wesley Wu wrote:
2010/8/20 Scott Ferguson f...@caucho.com:
I still think there's a chance for a Resin leak; I just don't understand
where it's coming from.
I think so. The same code worked fine (no leak) in Resin 4.0.5.
Do
Hi Scott,
I experienced ConnectionPool pool overflow issue again in resin 4.0.10.
Some time (probably 20 minutes) after resin started, resin reported:
[10-10-08 01:07:28.555] {server://*:6801-5523}
ConnectionPool[jdbc/mypool] pool overflow
[10-10-08 01:07:28.555] {server://*:6801-5109}
==beans.xml===
?xml version=1.0 encoding=UTF-8?
beans xmlns=http://java.sun.com/xml/ns/javaee;
xmlns:xsi=http://www.w3.org/2001/XMLSchema-instance;
xsi:schemaLocation=http://java.sun.com/xml/ns/javaee
http://java.sun.com/xml/ns/javaee/beans_1_0.xsd;
.
-Wesley
2010/11/2 Scott Ferguson f...@caucho.com:
I just checked with that example and it's working fine. Where are the
files/classes located, and how is JmsTemplate instantiated?
-- Scott
Wesley Wu wrote:
==beans.xml===
?xml version=1.0 encoding=UTF-8?
beans xmlns
Thanks, Scott. Would you give out a snapshot and I will do some stress
test on it.
-Wesley
2010/11/30 Scott Ferguson f...@caucho.com:
Thanks. I've fixed it for the next release (changing to ConcurrentHashMap)
- Scott
Wesley Wu wrote:
Or this method sync the wrong map?
private SetBean
Hi Scott,
Resin 4.0.14 log many warnings about no JMX ObjectName detected
[11-02-11 00:19:23.324] {resin-34} MemoryTenuredHealthCheck: WARNING:
MemoryTenuredHealthCheck[] has no JMX ObjectName detected
[11-02-11 00:19:23.324] {resin-34} MemoryPermGenHealthCheck: WARNING:
Hi Scott,
I found resin sometimes can't start ScheduledThreadPool, resulting in
@Inject
ExecutorService executorService
not work (executorService is null).
can't precisely reproduce it coz it happened randomly when server restarts.
affected resin version:
Resin 4.0.s100809
Resin 4.0.18
2.2 one.
I looked at the source of resin el implementation and found
java.el.CompositeELResolver did not implemented the
public Object invoke(ELContext context, Object base, Object method,
Class?[] paramTypes, Object[] params);
method which was added since EL 2.2.
any thoughts?
--
Wesley Wu
2.2 one.
I looked at the source of resin el implementation and found
java.el.CompositeELResolver did not implemented the
public Object invoke(ELContext context, Object base, Object method,
Class?[] paramTypes, Object[] params);
method which was added since EL 2.2.
any thoughts?
--
Wesley Wu
@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest
--
*Wesley Wu 吴萌野*
QQ:18990702
Mobile: 18601033886
Email: wesley...@yinyuetai.com wesley...@yinyuetai.com
工体北路8号三里屯SOHO办公D座1103 邮编100027
[image: U3%8Y%PN7N{6ZGQY1F)PI_P]
www.yinyuetai.com
* *
image001
http://caucho.com/resin-4.0/changes/changes.xtp
500 Servlet Exception
/var/www/hosts/www.caucho.com/webapps/resin-4.0/changes/changes.xtp:48:
`/li.' expected `' at `'. Closing tags must close immediately after
the tag name.
46: liconfig: app-oinf in cluster wasn't picked up properly (#5004)/li
64 matches
Mail list logo
