svn commit: r741579 - /tomcat/tc6.0.x/trunk/STATUS.txt
Author: jfclere Date: Fri Feb 6 14:47:22 2009 New Revision: 741579 URL: http://svn.apache.org/viewvc?rev=741579&view=rev Log: Propose my patch. Modified: tomcat/tc6.0.x/trunk/STATUS.txt Modified: tomcat/tc6.0.x/trunk/STATUS.txt URL: http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/STATUS.txt?rev=741579&r1=741578&r2=741579&view=diff == --- tomcat/tc6.0.x/trunk/STATUS.txt (original) +++ tomcat/tc6.0.x/trunk/STATUS.txt Fri Feb 6 14:47:22 2009 @@ -151,3 +151,8 @@ http://svn.apache.org/viewvc?rev=740684&view=rev +1: markt -1: + +* Make sure the 400 is returned to the browser. (like other connectors). + http://people.apache.org/~jfclere/patches/patch400.txt + +1: jfclere + -1: - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: svn repo layout for modules
Remy Maucherat wrote: On Fri, 2009-01-30 at 17:44 -0700, Filip Hanik - Dev Lists wrote: I would suggest https://svn.apache.org/repos/asf/tomcat/ - site - tc6.0.x - trunk - modules <-- ADD THIS - jdbc-pool - trunk - tags - branches - bayeux - trunk - tags - branches and move it away and out of (and get rid of) the tomcat/modules folder, and let tomcat be tomcat. My suggestion is to dump modules as much as possible, this is messy. JDBC should really be developed elsewhere beyond the experimental stage (Tomcat is not a DB project, thanks) and Bayeux should go to the main tree (if it works). That is a contradictory statement. Tomcat is not a DB project, nor is it a Bayeux project. So the reasoning above doesn't make any sense if you look at it that way. Tomcat is no longer the reference implementation either. So there is nothing that prevents us from providing value add components as long as we are willing to maintain it. ASF is community over code, its hard to grow the community if you never add any code. Tomcat JDBC would actually make TC 7 even more light weight since it is smaller than DBCP. Tomcat JDBC is not experimental either, its actually way more stable than expected. Run it through let me know if you find any bugs. So for modules, they are a good idea if we want to encourage people to contribute to more areas. as for the two modules, jdbc-pool and bayeux, I think they should still be modules if we want to keep the core light and by being modules let us fix bugs faster without releasing new versions of tomcat. So from a code perspective you are right, from a community perspective its not good at all to keep everything minimalistic. Especially since you mentioned you wouldn't involve yourself in trunk, then we are down by one resource. I would of course like for you to take back that statement and rejoin in the efforts, despite differences, the more is still the merrier So I suggest to keep the modules here, mostly to encourage new contributions to this project. Filip - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r741660 - /tomcat/tc6.0.x/trunk/STATUS.txt
Author: remm Date: Fri Feb 6 18:20:07 2009 New Revision: 741660 URL: http://svn.apache.org/viewvc?rev=741660&view=rev Log: - Votes. Modified: tomcat/tc6.0.x/trunk/STATUS.txt Modified: tomcat/tc6.0.x/trunk/STATUS.txt URL: http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/STATUS.txt?rev=741660&r1=741659&r2=741660&view=diff == --- tomcat/tc6.0.x/trunk/STATUS.txt (original) +++ tomcat/tc6.0.x/trunk/STATUS.txt Fri Feb 6 18:20:07 2009 @@ -131,28 +131,28 @@ * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=46564 Make page encoding test case insensitive - +1: markt + +1: markt, remm -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=46606 Make max depth configurable for WebDAV servlet http://svn.apache.org/viewvc?rev=740635&view=rev - +1: markt + +1: markt, remm -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=38197 Take account of jsp:attribute elements when naming tag pools http://svn.apache.org/viewvc?rev=740675&view=rev - +1: markt + +1: markt, remm -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=38553 Return 401 rather than 400 if client presents no certs for CLIENT-CERT auth http://svn.apache.org/viewvc?rev=740684&view=rev - +1: markt + +1: markt, remm -1: * Make sure the 400 is returned to the browser. (like other connectors). http://people.apache.org/~jfclere/patches/patch400.txt - +1: jfclere + +1: jfclere, remm -1: - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: svn repo layout for modules
On Fri, 2009-02-06 at 10:37 -0700, Filip Hanik - Dev Lists wrote: > ASF is community over code, its hard to grow the community if you never > add any code. > Tomcat JDBC would actually make TC 7 even more light weight since it is > smaller than DBCP. > Tomcat JDBC is not experimental either, its actually way more stable > than expected. Run it through let me know if you find any bugs. Tomcat does Comet, so Bayeux is really directly related. You can still move it if you think it is not appropriate, though. OTOH, it seems obvious a DB pool is purely a database project. There is a DB project at the ASF already, and there's also commons, where you can propose your project as DBCP.next. Tomcat then ships it, replacing DBCP, which results in the same value for users. I think you can add code which is Servlet or JSP related, rather than about anything which could get shipped with Tomcat. Rémy - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r741668 - /tomcat/tc6.0.x/trunk/STATUS.txt
Author: funkman Date: Fri Feb 6 18:44:13 2009 New Revision: 741668 URL: http://svn.apache.org/viewvc?rev=741668&view=rev Log: votes Modified: tomcat/tc6.0.x/trunk/STATUS.txt Modified: tomcat/tc6.0.x/trunk/STATUS.txt URL: http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/STATUS.txt?rev=741668&r1=741667&r2=741668&view=diff == --- tomcat/tc6.0.x/trunk/STATUS.txt (original) +++ tomcat/tc6.0.x/trunk/STATUS.txt Fri Feb 6 18:44:13 2009 @@ -126,33 +126,34 @@ * Align installer licenses with main license file http://svn.apache.org/viewvc?rev=734743&view=rev - +1: markt, fhanik + +1: markt, fhanik, funkman -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=46564 Make page encoding test case insensitive +1: markt, remm + +0: funkman: Where's the patch? -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=46606 Make max depth configurable for WebDAV servlet http://svn.apache.org/viewvc?rev=740635&view=rev - +1: markt, remm + +1: markt, remm, funkman -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=38197 Take account of jsp:attribute elements when naming tag pools http://svn.apache.org/viewvc?rev=740675&view=rev - +1: markt, remm + +1: markt, remm, funkman -1: * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=38553 Return 401 rather than 400 if client presents no certs for CLIENT-CERT auth http://svn.apache.org/viewvc?rev=740684&view=rev - +1: markt, remm + +1: markt, remm, funkman -1: * Make sure the 400 is returned to the browser. (like other connectors). http://people.apache.org/~jfclere/patches/patch400.txt - +1: jfclere, remm + +1: jfclere, remm, funkman -1: - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r741747 - in /tomcat/trunk/modules/jdbc-pool: java/org/apache/tomcat/jdbc/pool/ java/org/apache/tomcat/jdbc/pool/jmx/ test/org/apache/tomcat/jdbc/test/
Author: fhanik
Date: Fri Feb 6 21:49:36 2009
New Revision: 741747
URL: http://svn.apache.org/viewvc?rev=741747&view=rev
Log:
Make sure thread starvation doesn't exist during reclamation of connections
Added:
tomcat/trunk/modules/jdbc-pool/test/org/apache/tomcat/jdbc/test/StarvationTest.java
(with props)
Modified:
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/ConnectionPool.java
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/PooledConnection.java
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/jmx/ConnectionPool.java
Modified:
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/ConnectionPool.java
URL:
http://svn.apache.org/viewvc/tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/ConnectionPool.java?rev=741747&r1=741746&r2=741747&view=diff
==
---
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/ConnectionPool.java
(original)
+++
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/ConnectionPool.java
Fri Feb 6 21:49:36 2009
@@ -98,7 +98,12 @@
* reference to mbean
*/
protected org.apache.tomcat.jdbc.pool.jmx.ConnectionPool jmxPool = null;
-
+
+/**
+ * counter to track how many threads are waiting for a connection
+ */
+protected AtomicInteger waitcount = new AtomicInteger(0);
+
//===
// PUBLIC METHODS
//===
@@ -142,6 +147,14 @@
public String getName() {
return getPoolProperties().getPoolName();
}
+
+/**
+ * Return the number of threads waiting for a connection
+ * @return number of threads waiting for a connection
+ */
+public int getWaitCount() {
+return waitcount.get();
+}
/**
* Returns the pool properties associated with this connection pool
@@ -397,6 +410,9 @@
jmxPool.notify(org.apache.tomcat.jdbc.pool.jmx.ConnectionPool.NOTIFY_ABANDON,
trace);
}
con.abandon();
+//we've asynchronously reduced the number of connections
+//we could have threads stuck in idle.poll(timeout) that will
never be notified
+if (waitcount.get()>0) idle.offer(new
PooledConnection(poolProperties,this));
} finally {
con.unlock();
}
@@ -457,11 +473,17 @@
maxWait =
(getPoolProperties().getMaxWait()<=0)?Long.MAX_VALUE:getPoolProperties().getMaxWait();
}
long timetowait = Math.max(0, maxWait -
(System.currentTimeMillis() - now));
+waitcount.incrementAndGet();
try {
//retrieve an existing connection
con = idle.poll(timetowait, TimeUnit.MILLISECONDS);
} catch (InterruptedException ex) {
-Thread.interrupted();
+Thread.interrupted();//clear the flag, and bail out
+SQLException sx = new SQLException("Pool wait interrupted.");
+sx.initCause(ex);
+throw sx;
+} finally {
+waitcount.decrementAndGet();
}
if (maxWait==0 && con == null) { //no wait, return one if we have
one
throw new SQLException("[" +
Thread.currentThread().getName()+"] " +
@@ -529,6 +551,9 @@
boolean setToNull = false;
try {
con.lock();
+if (!con.isDiscarded() && !con.isInitialized()) {
+con.connect();
+}
if ((!con.isDiscarded()) &&
con.validate(PooledConnection.VALIDATE_BORROW)) {
//set the timestamp
con.setTimestamp(now);
@@ -647,7 +672,6 @@
if ((now - time) > con.getAbandonTimeout()) {
busy.remove(con);
abandon(con);
-release(con);
setToNull = true;
} else {
//do nothing
Modified:
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/PooledConnection.java
URL:
http://svn.apache.org/viewvc/tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/PooledConnection.java?rev=741747&r1=741746&r2=741747&view=diff
==
---
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/PooledConnection.java
(original)
+++
tomcat/trunk/modules/jdbc-pool/java/org/apache/tomcat/jdbc/pool/PooledConnection.java
Fri Feb 6 21:49:36 2009
@@ -53,8 +53,8 @@
protected ConnectionPool parent;
protected WeakReference handler = null;
-
-public PooledConnection(PoolProperties prop, ConnectionPool parent) throws
SQLException {
+
+publ
Re: svn repo layout for modules
Remy Maucherat wrote: On Fri, 2009-02-06 at 10:37 -0700, Filip Hanik - Dev Lists wrote: ASF is community over code, its hard to grow the community if you never add any code. Tomcat JDBC would actually make TC 7 even more light weight since it is smaller than DBCP. Tomcat JDBC is not experimental either, its actually way more stable than expected. Run it through let me know if you find any bugs. Tomcat does Comet, so Bayeux is really directly related. You can still move it if you think it is not appropriate, though. OTOH, it seems obvious a DB pool is purely a database project. There is a DB project at the ASF already, and there's also commons, where you can propose your project as DBCP.next. Tomcat then ships it, replacing DBCP, which results in the same value for users. that discussion does exist if you look back into the archives what has changed since then? Filip I think you can add code which is Servlet or JSP related, rather than about anything which could get shipped with Tomcat. Rémy - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
DO NOT REPLY [Bug 46666] Why is disableUploadTimeout and KeepAliveTimeout correlated in Http11 Connector?
https://issues.apache.org/bugzilla/show_bug.cgi?id=4 Doug Drechsel changed: What|Removed |Added Status|NEW |NEEDINFO -- Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
DO NOT REPLY [Bug 46666] Why is disableUploadTimeout and KeepAliveTimeout correlated in Http11 Connector?
https://issues.apache.org/bugzilla/show_bug.cgi?id=4 Doug Drechsel changed: What|Removed |Added Status|NEEDINFO|NEW -- Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: svn repo layout for modules
On Fri, 2009-02-06 at 14:59 -0700, Filip Hanik - Dev Lists wrote: > that discussion does exist if you look back into the archives > what has changed since then? The proposed module layout looks so-so to me, and apparently others feel the same, so a solution is to remove the modules. Rémy - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: svn repo layout for modules
On Fri, Feb 6, 2009 at 3:07 PM, Remy Maucherat wrote: > On Fri, 2009-02-06 at 14:59 -0700, Filip Hanik - Dev Lists wrote: > > that discussion does exist if you look back into the archives > > what has changed since then? > > The proposed module layout looks so-so to me, and apparently others feel > the same, so a solution is to remove the modules. Can we just leave the modules as they are - it really doesn't matter that much. Tagging is possible ( just tag the whole tree ), release is possible. The question of scope remains - but I see no problem with having a tomcat-specific jdbc pool or other components ( like we have juli and many other things that could fit commons better ), it it is maintained ( which makes it quite simple - if code is maintained mostly by tomcat, why not leave it in tomcat ? ) I'm going to push the lite 'module' as well - modules dir seems a good middle ground. Costin
svn commit: r741836 - /tomcat/tc6.0.x/trunk/STATUS.txt
Author: fhanik Date: Sat Feb 7 03:22:15 2009 New Revision: 741836 URL: http://svn.apache.org/viewvc?rev=741836&view=rev Log: votes Modified: tomcat/tc6.0.x/trunk/STATUS.txt Modified: tomcat/tc6.0.x/trunk/STATUS.txt URL: http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/STATUS.txt?rev=741836&r1=741835&r2=741836&view=diff == --- tomcat/tc6.0.x/trunk/STATUS.txt (original) +++ tomcat/tc6.0.x/trunk/STATUS.txt Sat Feb 7 03:22:15 2009 @@ -150,10 +150,10 @@ * Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=38553 Return 401 rather than 400 if client presents no certs for CLIENT-CERT auth http://svn.apache.org/viewvc?rev=740684&view=rev - +1: markt, remm, funkman + +1: markt, remm, funkman, fhanik -1: * Make sure the 400 is returned to the browser. (like other connectors). http://people.apache.org/~jfclere/patches/patch400.txt - +1: jfclere, remm, funkman + +1: jfclere, remm, funkman, fhanik -1: - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
svn commit: r741843 - /tomcat/tc6.0.x/trunk/STATUS.txt
Author: billbarker Date: Sat Feb 7 05:18:37 2009 New Revision: 741843 URL: http://svn.apache.org/viewvc?rev=741843&view=rev Log: voting Modified: tomcat/tc6.0.x/trunk/STATUS.txt Modified: tomcat/tc6.0.x/trunk/STATUS.txt URL: http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/STATUS.txt?rev=741843&r1=741842&r2=741843&view=diff == --- tomcat/tc6.0.x/trunk/STATUS.txt (original) +++ tomcat/tc6.0.x/trunk/STATUS.txt Sat Feb 7 05:18:37 2009 @@ -151,9 +151,12 @@ Return 401 rather than 400 if client presents no certs for CLIENT-CERT auth http://svn.apache.org/viewvc?rev=740684&view=rev +1: markt, remm, funkman, fhanik + 0: billbarker With the JIO connector, this won't do anything since the connection is already dropped + Haven't checked what the other ones will do -1: * Make sure the 400 is returned to the browser. (like other connectors). http://people.apache.org/~jfclere/patches/patch400.txt +1: jfclere, remm, funkman, fhanik - -1: + -1: billbarker Even though I'm already out voted, this needs to go through the ActionHook to + prevent confusing Apache with multiple END_RESPONSE messages - To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
Re: svn repo layout for modules
What we and user need, is a real plugin and tool contribution area. +1 to start the modules plugin system to create more value of the tomcat project regards, Peter Am 07.02.2009 um 01:18 schrieb Costin Manolache: On Fri, Feb 6, 2009 at 3:07 PM, Remy Maucherat wrote: On Fri, 2009-02-06 at 14:59 -0700, Filip Hanik - Dev Lists wrote: that discussion does exist if you look back into the archives what has changed since then? The proposed module layout looks so-so to me, and apparently others feel the same, so a solution is to remove the modules. Can we just leave the modules as they are - it really doesn't matter that much. Tagging is possible ( just tag the whole tree ), release is possible. The question of scope remains - but I see no problem with having a tomcat-specific jdbc pool or other components ( like we have juli and many other things that could fit commons better ), it it is maintained ( which makes it quite simple - if code is maintained mostly by tomcat, why not leave it in tomcat ? ) I'm going to push the lite 'module' as well - modules dir seems a good middle ground. Costin
