WordPress mod_rewrite rules taking over mod_status and mod_info

After moving Rule of Tech to a new server and setting up monitoring I noticed that server-status and server-info Apache modules weren’t working as expected. As usual a little bit of Googling solved this problem.

The problem was that the .htaccess rules in WordPress were taking over non-existing server-info and server-status urls given in Apache’s config and were returning a page not found error. The rewrite rules by WordPress were setup to handle all the permalinks on the site and for any non-existing file send it to index.php. It really wasn’t a WordPress problem and should happen with any application that uses the same type of catch-all rewrite rules to handle all the urls inside the application.

The solution was to specifically add a rewrite rule to not have the server-status and server-info urls processed by adding a rule like: RewriteCond %{REQUEST_URI} !=/server-status. The other way is to stop the rewriting process when the urls are found by adding a rule like: RewriteRule ^(server-info|server-status) - [L].

The WordPress rewrite rules should look like this:

# BEGIN WordPress
<IfModule mod_rewrite.c>
RewriteEngine On
RewriteBase /
# server info and status
RewriteRule ^(server-info|server-status) - [L]
# RewriteCond %{REQUEST_URI} !=/server-status
# /server info and status
RewriteCond %{REQUEST_FILENAME} -f 
RewriteCond %{REQUEST_FILENAME} -d
RewriteRule . index.php [L]
</IfModule>
# END WordPress

Installing Apache Tomcat 6 on CentOS

CentOS is great substitute for Red Hat Enterprise Linux but is missing some useful packages like Apache Tomcat 6. Installing Apache Tomcat 6 on CentOS 5 from gzip-package is fairly easy. The following guide is at least for CentOS 5.4.

Pre-Requirements
First you need to install Sun JDK and you can follow the instructions given in Installing Sun JDK 1.6 on CentOS

After Java is on place it’s time to get ready for Tomcat.

Download Apache Ant and Tomcat

  1. Download apache-ant and apache-tomcat -packages.
  2. Extract those packages to /opt/
    • #[root@srv ~]# cd /opt
      # tar -xzf apache-tomcat-6.0.26.tar.gz
      # tar -xzf apache-ant-1.7.1-bin.tar.gz
      
  3. Create a symbolic link for Ant
    • # ln -s /opt/apache-ant-1.7.1/bin/ant /usr/bin/
      

Create start script

  1. Create a tomcat user so that we don’t need root privileges for Tomcat
    • # useradd -d /opt/apache-tomcat-6.0.26/ tomcat
      
  2. Create start script to /etc/init.d for starting and stopping Tomcat
    • #  vim /etc/init.d/tomcat
      
  3. The script is (via Build a safe cage for Tomcat)
    • 1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
      13
      14
      15
      16
      17
      18
      19
      20
      21
      22
      23
      24
      25
      26
      27
      28
      29
      30
      31
      32
      33
      34
      35
      36
      37
      38
      39
      40
      41
      42
      43
      44
      45
      46
      47
      48
      49
      50
      51
      52
      53
      54
      55
      56
      57
      58
      59
      60
      
      #!/bin/bash
      #
      # tomcat       Starts Tomcat Java server.
      #
      #
      # chkconfig: 345 88 12
      # description: Tomcat is the server for 
      # Java servlet applications.
      ### BEGIN INIT INFO
      # Provides: $tomcat
      ### END INIT INFO
       
      JAVA_HOME=/usr/java/jdk1.6.0_18
      export JAVA_HOME
      TOMCAT_HOME=/opt/apache-tomcat-6.0.26/bin
      START_TOMCAT=/opt/apache-tomcat-6.0.26/bin/startup.sh
      STOP_TOMCAT=/opt/apache-tomcat-6.0.26/bin/shutdown.sh
       
      # Source function library.
      . /etc/init.d/functions
       
      [ -f $START_TOMCAT ] || exit 0
      [ -f $STOP_TOMCAT ] || exit 0
       
      RETVAL=0
       
      umask 077
       
      start() {
              echo -n $"Starting Tomcat Java server: "
              daemon su -c $START_TOMCAT tomcat
              echo
              return $RETVAL
      }
      stop() {
              echo -n $"Shutting down Tomcat Java server: "
              daemon su -c $STOP_TOMCAT tomcat
              echo
              return $RETVAL
      }
      restart() {
              stop
              start
      }
      case "$1" in
        start)
              start
              ;;
        stop)
              stop
              ;;
        restart|reload)
              restart
              ;;
        *)
              echo $"Usage: $0 {start|stop|restart}"
              exit 1
      esac
       
      exit $?
  4. Give executable rights for that script
    • # chmod 755 /etc/init.d/tomcat
      
  5. Add the script to CentOS services
    • # chkconfig --add tomcat
      
  6. Check the changes
    • # chkconfig --level 234 tomcat on
      # chkconfig --list tomcat
      
      tomcat 0:off 1:off 2:on 3:on 4:on 5:off 6:off
      
  7. You should see that the service uses levels 2, 3 and 4
  8. Test that the script is working and it gives no errors
    • # service tomcat start
      # service tomcat stop
      
  9. Everythings ready

Installing Sun JDK 1.6 on CentOS

CentOS doesn’t have a package for Sun JDK so it has to installed manually. It’s fairly easy but there are some steps to do that. This guide has been tested on CentOS 5.4 x64_86.

Step 1. Initial setup for building RPM
-!- Do this with a non-root user

  1. Create ~/.rpmmacros
    • $ vim ~/.rpmmacros
      %_topdir /home//rpmbuild
      %_tmppath %{_topdir}/tmp
      
  2. Create needed folders:
    • $ mkdir -p ~/rpmbuild/{SOURCES,SRPMS,SPECS,RPMS,tmp,BUILD}
      
  3. Build environment needs to be complete. Some needed packages are:
    • $ sudo yum install -y rpm-build gcc gcc-c++ redhat-rpm-config
      

Step 2. Installing your favorite JDK

  1. Download Sun JDK 1.6 update 14 from Sun Java download or the Sun JDK archive.
    • Choose the correct platform (for me it’s Linux x64) and download jdk-6u18-linux-x64-rpm.bin
  2. Give it executable rights: $ chmod 755 jdk-6u18-linux-x64-rpm.bin
  3. Run the binary to extract it into RPM form: $ ./jdk-6u18-linux-x64-rpm.bin
  4. Install it:
    • $ sudo rpm -Uvh jdk-6u18-linux-amd64.rpm
      
  5. Log out and in again to make the changes in the paths take effect
  6. Check the install
    • $ java -version
      java version "1.6.0_18"
      Java(TM) SE Runtime Environment (build 1.6.0_18-b07)
      Java HotSpot(TM) 64-Bit Server VM (build 16.0-b13, mixed mode)
      
  7. Java is now installed on /usr/bin/java

Using RichFaces 3 dataScroller and dataTable -components

RichFaces provides some nice AJAX-components for Java Server Faces but the documentation and examples could be better. RichFaces has great documentation compared to some other frameworks but it could be better with adding a little bit of real world and down to earth examples. So here is one example of using RichFaces dataScroller and dataTable -components with custom CSS-styling, backingBean and JSF-page snippets using Richfaces 3.3.2.SR1 and JSF 1.2_12.

Using RichFaces dataScroller and dataTable components has a big negative property: they work nicely if the amount of data is small but when the row count reaches to thousands they become sluggish or stop working. The rich:dataScroller needs the complete datamodel being loaded into memory and only displays a part of it. Not very efficient if the rowcount exceeds 1000 or so.

Anyways here is some real world example. The icons used in the examples for dataScroller are from Crystal Project Icons.

JSF-page

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
<h:form id="myForm">
    <rich:dataScroller styleClass="dataScroller" id="scroller" 
        for="resultTable" maxPages="15" fastStep="3"
        renderIfSinglePage="false" immediate="false" 
        binding="#{backingBean.scroller}" page="#{backingBean.scrollerPage}">
        <f:facet name="first" >
            <h:graphicImage id="firstImage" styleClass="scroller" 
            url="images/crystal/tab_first.png" alt="first"/>
        </f:facet>
        <f:facet name="last">
            <h:graphicImage id="lastImage" styleClass="scroller" 
                url="images/crystal/tab_last.png" alt="last"/>
        </f:facet>
        <f:facet name="previous">
            <h:graphicImage id="prevImage" styleClass="scroller" 
                url="images/crystal/tab_left.png" alt="previous"/>
            </f:facet>
        <f:facet name="next">
            <h:graphicImage id="nextImage" styleClass="scroller" 
                url="images/crystal/tab_right.png" alt="previous"/>
        </f:facet>
        <f:facet name="fastforward">
            <h:graphicImage id="ffImage" styleClass="scroller" 
                url="images/crystal/tab_fastf.png" alt="next"/>
        </f:facet>
        <f:facet name="fastrewind">
            <h:graphicImage id="frImage" styleClass="scroller" 
                url="images/crystal/tab_fastr.png" alt="next"/>
        </f:facet>
        <f:facet name="first_disabled" >
            <h:graphicImage id="firstImage_d" styleClass="scroller" 
                url="images/crystal/tab_first.png" alt="first"/>
        </f:facet>
        <f:facet name="last_disabled">
            <h:graphicImage id="lastImage_d" styleClass="scroller" 
                url="images/crystal/tab_last.png" alt="last"/>
        </f:facet>
        <f:facet name="previous_disabled">
            <h:graphicImage id="prevImage_d" styleClass="scroller" 
                url="images/crystal/tab_left.png" alt="previous"/>
        </f:facet>
        <f:facet name="next_disabled">
            <h:graphicImage id="nextImage_d" styleClass="scroller" 
                url="images/crystal/tab_right.png" alt="next"/>
        </f:facet>
        <f:facet name="fastforward_disabled">
            <h:graphicImage id="ffImage_d" styleClass="scroller" 
                url="images/crystal/tab_fastf.png" alt="next"/>
        </f:facet>
        <f:facet name="fastrewind_disabled">
            <h:graphicImage id="frImage_d" styleClass="scroller" 
                url="images/crystal/tab_fastr.png" alt="previous"/>
        </f:facet>
        <f:facet name="controlsSeparator">
            <h:outputText id="sep" value=" " />
        </f:facet>
</rich:dataScroller>
 
<rich:dataTable styleClass="resultTable" id="resultTable" 
    rows="10" rowClasses=",odd"  columnClasses="col"  
    value="#{backingBean.resultList}" binding="#{backingBean.resultData}" 
    var="h" sortMode="multi">
    <rich:column sortBy="#{h.desc}">
        <f:facet name="header">
            <h:outputText value="description" />
        </f:facet>
        <h:commandLink value="#{h.desc}"
            action="#{backingBean.showRowData}">
            <f:param name="selectedRow" value="#{h.desc" />
        </h:commandLink>
    </rich:column>
    <rich:column sortBy="#{h.value}">
        <f:facet name="header">
            <h:outputText value="value" />
        </f:facet>
        <h:outputText value="#{h.value}" />
    </rich:column>
</rich:dataTable>
</h:form>

Backing Bean

Create some variables for dataScroller and getters and setters for them:

1
2
3
4
5
6
7
8
// RichFaces dataScroller variables
private HtmlDatascroller scroller = new HtmlDatascroller();
private String scrollerPage = "";
 
// Getting the clicked row's data
public String showRowdata() {
  MyDataModel current = (myDataModel) getResultData().getRowData();
}

CSS styling

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
/* =RichFaces DataScroller
----------------------------------------------- */
.rich-datascr {font-size: 1.1em;border: 0;}
.rich-table-cell {font-size: 1.0em;}
.rich-table-sortable-header {font-size: 1.1em;font-weight: bold;}
td.rich-datascr-button {background-color: #fff;border: 0px solid #ccc;text-decoration: none;}
td.rich-datascr-button-dsbld {background-color: #fff;}
.rich-datascr-ctrls-separator {padding-right: 5px;}
.rich-dtascroller-table {background: #fff;border: 0;}
.scroller {display: block;background-color: #fff;border: 1px solid #ccc;padding: 3px 3px;margin: 0px 5px 5px 5px;text-decoration: none;}
.scroller:hover {background-color: #eee;}
td.rich-datascr-button-dsbld .scroller {background-color: #eee;}
td.rich-datascr-inact {font-size: 1.2em;color: #000;border: 0;}
td.rich-datascr-inact:hover {text-decoration: underline;}
td.rich-datascr-act {font-size: 1.2em;text-decoration: underline;}
td.rich-datascr-act {border: 0;font-weight: bold;}

Selecting All rows with JavaScript

Add to the JSF-page a new column which has the checkbox. We are using JavaScript to loop through the input fields which are after :tu -ending id-field.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
<rich:column>
  <f:facet name="header">
    <h:panelGroup layout="block">
      <script type="text/javascript">
        //<![CDATA[
        // RichFaces datatable select all -checkbox
        function checkAllCheckboxesInTable( inputId, state ){
          var commonIdPart = inputId.substr(0, inputId.lastIndexOf(':'));
          var tableId = commonIdPart + ':tu'
          var tableElement = document.getElementById( tableId );
          var inputs = tableElement.getElementsByTagName('input');
          for (var i = 0; i <= inputs.length; i++){
            var input = inputs[i];
            if (input != undefined) {
              if( input.getAttribute('type') == 'checkbox' && state){
                input.setAttribute('checked', state);
              } else{
                input.setAttribute('checked', false);
                input.removeAttribute('checked');
              }
            }
          }
        }
        //]]>
      </script>
      <h:selectBooleanCheckbox id="t0" onclick="checkAllCheckboxesInTable( this.id, this.checked );">
        <a4j:support event="onchange" reRender="resultTable"/>
      </h:selectBooleanCheckbox>
    </h:panelGroup>
  </f:facet>
  <h:selectBooleanCheckbox id="t1" value="#{h.selected}" />
</rich:column>

Selecting All rows in backing bean

You can also check all the checkboxes from the backingBean but it has problems with table ordering and when the lists order changes the selection goes wrong.

Add to the JSF-page a new column:

1
2
3
4
5
6
7
8
<rich:column>
  <f:facet name="header">
    <h:selectBooleanCheckbox id="t0" value="#{backingBean.selectedAll}" onclick="this.blur()">
        <a4j:support event="onchange" actionListener="#{backingBean.selectAll}" reRender="resultTable, t0, t1"/>
      </h:selectBooleanCheckbox>
    </f:facet>
  <h:selectBooleanCheckbox id="t1" value="#{h.selected}" />
</rich:column>

Make a new method to your backingBean:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
public void selectAll(ActionEvent event) {
	logger.info("*** backingBean.selectAll(): " + scrollerPage + " ***");
 
	// get the current scroller page
	int page = Integer.valueOf(scrollerPage).intValue();
	if (page != 0) {
	    page = page - 1;
	}
	int start = page * 10;
	int stop = (page * 10) + 10;
	if (stop > getResultList().size()) {
	    stop = getResultList().size();
	}
	logger.debug("> page: " + page + "; start: " + start + "; stop: " + stop);
 
	// check the boxes on the active page
	for (int i = start; i < stop; i++) {
	    logger.debug("> valitaan: " + i + "; " + selectedAll);
	    getResultList().get(i).setSelected(selectedAll);
	}
}

Eclipse: Class file name must end with .class exception in search

Eclipse is nice IDE but it has it’s own problems. This time the Java Search and Open Type -search produced an error saying “Class file name must end with .class”. Very helpfull. Fortunately almost all the answers in the world can be found in the Internet and so with a quick googling the solution to this annoying problem was found on Stack Overflow.

I had already tried Project -> Clean… and closing Eclipse, deleting all the built class files and restarting Eclipse to no avail as was the original question author. The right answer lies in deleting the corrupted search index which is explained in Eclipse bug’s #269820 comment.

How to delete the search index:

  1. Close Eclipse
  2. Delete workspace/.metadata/.plugins/org.eclipse.jdt.core/*.index
  3. Delete workspace/.metadata/.plugins/org.eclipse.jdt.core/savedIndexNames.txt
  4. Start Eclipse again

This fixed the issue for me.

Redirect HTTP and HTTPS traffic to Tomcat’s ports

Apache Tomcat likes with default settings to listen to requests on 8080 and 8443 ports but it is more enjoyable to use the more common 80 and 443 ports for HTTP and HTTPS traffic. This way the user don’t have to put those pesky port numbers after the address. Of course you could just tell Tomcat to listen to those ports but it has some negative sides: hassle with the startup and running Tomcat as root.

Luckily it is easy to tell the system to redirect the traffic from some port to other. Just define some new xinetd services in /etc/xinetd.d/tomcat.

# vim /etc/xinetd.d/tomcat
 
# Redirects any requests on port 80 to port 8080 (where Tomcat is listening)
service tomcat-http
{
        disable                 = no
        flags                   = REUSE
        wait                    = no
        user                    = root
        socket_type         = stream
        protocol                = tcp
        port                    = 80
        redirect                = localhost 8080
        log_on_success  -= PID HOST DURATION EXIT
 
        #per_source = UNLIMITED
        #instances = UNLIMITED
}
 
# Redirects any requests on port 443 to port 8443 (where Tomcat is listening)
service tomcat-https
{
        disable                 = no
        flags                   = REUSE
        wait                    = no
        user                    = root
        socket_type         = stream
        protocol                = tcp
        port                    = 443
        redirect                = localhost 8443
        log_on_success  -= PID HOST DURATION EXIT
 
        #per_source = UNLIMITED
        #instances = UNLIMITED
}

(via Securing Linux for Java services: The port dilemma)

Xinetd puts a connection limit per source IP, by default and this causes the service to become unresponsive when there are dozens of queries a second. You see the following kind of line in your messages log file: “xinetd[2049]: FAIL: tomcat-https per_source_limit from=123.456.789.123”. To correct this, uncomment the per_source and instances lines in your xinet.d file and restart it.

Also add those xinetd services to /etc/services.

# vim /etc/services
http        80/tcp     www www-http tomcat-http # WorldWideWeb http
http        80/udp     www www-http tomcat-http # WorldWideWeb HTTP
http        443/tcp    tomcat-https # WorldWideWeb HTTPS
http        443/udp    tomcat-https # WorldWideWeb HTTPS

And now just restart the xinetd and admire how your traffic is redirected to Tomcat’s ports.

# service xinetd restart

Force everything to transmit through HTTPS
If you also want to redirect all HTTP traffic to HTTPS you can add the following section to you Tomcat web.xml:

<web-resource-collection>
    <web-resource-name>Protected Context</web-resource-name>
    <url-pattern>/*</url-pattern>
</web-resource-collection>
<!-- auth-constraint goes here if you requre authentication -->
<user-data-constraint>
    <transport-guarantee>CONFIDENTIAL</transport-guarantee>
</user-data-constraint>

If you are using this redirection of all traffic to HTTPS with JIRA and want to attachments working also with Internet Explorer then you must add the following to your jira.xml (f. ex. /opt/tomcat/conf/Catalina/localhost/jira.xml). This is a Internet Explorer bug, for more information see http://jira.atlassian.com/browse/JRA-8179.

<Context ...>
...
<!-- for IE bug, see http://jira.atlassian.com/browse/JRA-8179-->
<Valve className="org.apache.catalina.authenticator.NonLoginAuthenticator"
disableProxyCaching="false" />
...
</Context>

Keeping up with the time in Xen

A simple tip for keeping up with the time in Xen when your domU isn’t syncing the clock frequently enough and your clock is whatever. Of course you can sync your clock with ntpdate but first you need to set the domain (domU) to run its wallclock independently from Xen.

Simply say in the command line of domU:
echo 1 > /proc/sys/xen/independent_wallclock

Or if you have noclobber on:
echo 1 >! /proc/sys/xen/independent_wallclock

To reenable tracking of Xen wallclock:
echo 0 > /proc/sys/xen/independent_wallclock

And to keep the setting between reboots, just add it to the /etc/sysctl.conf:
xen.independent_wallclock = 1

Although, it is better if the Xen dom0 syncs the clock frequently so all the domains (domU) don’t need to do it by themselves and thus wasting resources. Why do things multiple times when it could be done just once.

Samba performance problem after kernel update

Samba is a nice service to provide storage space through networks and it is relatively easy to set up. It’s not as fast as using NFS between Linux hosts but sometimes you don’t have that possibility. Sometimes there also might be quite confusing problems like I happened to notice.

My Samba service had worked fine for a long time but after a kernel update and a reboot, the performance was horrible. It was really slow to do anything. Fortunately I wasn’t the only one to come by with this problem and there was a topic Samba Performance Problem Due to Changing Linux Kernel in Samba Performance Tuning -guide which provided a solution to my problem. Just restarted the network interface and the performance was as good as before.

Earlier this month Samba Team Received Microsoft Protocol Docs so maybe in the near future we will get better and faster software.

The Protocol Freedom Information Foundation (PFIF), signed an agreement with Microsoft to receive the protocol documentation needed to fully interoperate with the Microsoft Windows workgroup server products and to make them available to Free Software projects such as Samba.

Microsoft was required to make this information available to competitors as part of the European Commission March 24th 2004 Decision in the antitrust lawsuit, after losing their appeal against that decision on September 17th 2007.
Samba Team

WordPress and problem with language files (.mo)

While testing the upcoming WordPress 2.3 I noticed that I had a problem with he localization. Defining the language with WPLANG had no effects and all the text were in English as they should have been in Finnish. I installed another instance of WordPress 2.2.3 and the problem was still there. Somewhat confusing because I had a working WordPress instance with the same PHP-version (php-5.2.4_pre200708051230-r2). The only difference was that the other box was 32-bit system and the other 64-bit. Although at that time I didn’t know what to look for.

After some googling I found a Spanish-blog which explained the “bug” and provided a solution. Unfortunately I don’t speak Spanish but the solution was clear with pieces of code to show the way. After I knew what to look for I found some discussions about this topic on WordPress Trac and an entry 2 weeks ago with a patch gettext-64-without-bitwise-ops.diff seems to get the thing fixed.

The problem is in the PHP-gettext which is used for translating strings. The file is located at wp-includes/gettext.php and in that file the problem is created by fixing a bug in PHP 5.0.2 on 64-bit systems. The counterfix is luckily an easy one and you just have to remove some bitwise operators.

The solution with problematic line on red and the fix on green:

115
116
117
118
119
120
121
122
123
124
wp-includes/gettext.php
$this-&gt;STREAM = $Reader;
$magic = $this-&gt;readint();
117c117
< if ($magic == ($MAGIC1 &amp; 0xFFFFFFFF) || $magic == ($MAGIC3 &amp; 0xFFFFFFFF)) { // to make sure it works for 64-bit platforms
---
> if ($magic == $MAGIC1 || $magic == $MAGIC3) { // to make sure it works for 64-bit platforms
>	$this-&gt;BYTEORDER = 0;
> } elseif ($magic == ($MAGIC2 &amp; 0xFFFFFFFF)) {
>	$this-&gt;BYTEORDER = 1;

The search engines <3

Update, 2008-02-05:
With WordPress 2.3.3 this bug is finally closed with some other minor bug fixes and security fix for xmlrpc.php.

Apache can't start after updating expat to 2.0.1

After updating to expat-2.0.1, it tells you to run revdep-rebuild -X --library=libexpat.so.0 but even after that apache2 couldn’t be started. It says that /usr/sbin/apache2 can’t find shared library libexpat.so.0. After some googling I found a message in Google Groups which deals with the problem.

The problem is that /usr/lib/libaprutil-0.so.0.9.12 which belongs to apr-util is still broken. But how can that be broken when you just ran the revdep-rebuild which should have fixed the problem. The reason is that by executing eix apr-util, you see that there are two versions of apr-util installed.

After re-emergeing the apr-util-0.9.12-r1, apache2 starts normally.

The problem is actually caused by the “-X” option of revdep-rebuild command, which emerges the best (currrently is the latest stable) packages available. In this apr-util case, it will emerge apr-util-1.2.8.

So for slotted packages, don’t use “-X” option for revdep-rebuild command.
Shaochun Wang