refer to:
https://support.oracle.com/epmos/faces/ui/km/SearchDocDisplay.jspx?returnToSrId=&_afrLoop=233778927223699&srnum=&type=DOCUMENT&id=435058.1&displayIndex=1&_afrWindowMode=0&_adf.ctrl-state=7nm491gg4_154
This error condition is normally seen when the invalidation message has not arrived at any WebCache (eg, the name and port used is incorrect or unresponsive). There are three possible reasons:-
(i) The Host Name as stored in the Portal metadata is wrong and cannot be found.
This problem error will mean and invalidation message will fail with WWC-00019 after a TCP timeout period, typically 15 seconds.
(ii) The Port as stored in the Portal metadata is wrong and cannot be reached.
This problem error will mean and invalidation message will fail with WWC-00019 immediately
(iii) The Host Name and Port as stored in the Portal metadata is correct but cannot be transmitted for some network reason.
This problem error will mean and invalidation message will fail with WWC-00019 immediately.
In all 3 cases, the error seen in the user's browser will be:-
Error: Could not open web cache connection. (WWC-40019)
Solution
Testing the Network
Before trying to update the Portal metadata, its important to validate the hostname and port used do refer to a functioning WebCache server. Some simple network tests can be used for this validation.
These test MUST be carried out on ALL Portal Database Servers and not the Portal Midtier nor the Client PC.
1. ping <hostname> - this needs to return success
2. telnet <hostname> <invalidation port> - this needs to return immediately with a result like the text below:
$ telnet portal.uk.oracle.com 9401
Trying 10.11.12.13...
Connected to portal.uk.oracle.com (10.11.12.13).
Escape character is '^]'.
Trying 10.11.12.13...
Connected to portal.uk.oracle.com (10.11.12.13).
Escape character is '^]'.
3. Login to the metadata database as the PORTAL user or the SYSTEM user and issue the following SQL statement:
select utl_http.request('http://<hostname>:<invalidation port>') from sys.dual;
If the hostname:port combination is good, you should get test like the example below:
UTL_HTTP.REQUEST('HTTP://<hostname>:<invalidation port>')
--------------------------------------------------------------------------------
<HTML><HEAD><TITLE>Unauthorized</TITLE></HEAD><BODY>This document is protected.
You must send the<BR>proper authorization information to access it.</BODY></HTML>
--------------------------------------------------------------------------------
<HTML><HEAD><TITLE>Unauthorized</TITLE></HEAD><BODY>This document is protected.
You must send the<BR>proper authorization information to access it.</BODY></HTML>
If the hostname:port combination is bad, the text may show:
ERROR at line 1:
ORA-06512: at "SYS.UTL_HTTP", line 1577
ORA-06512: at "SYS.UTL_HTTP", line 1577
ORA-12545: Connect failed because target host or object does not exist
ORA-06512: at line 1
Until these tests succeed, it it not worth trying to correct the Portal metadata.
Note:
The most common reason for these tests to fail is the local hosts file is setup incorrectly.
The most common reason for these tests to fail is the local hosts file is setup incorrectly.
Changing the Portal MetaData
11g
For all these error conditions, the Portal will carry on working, you will still be able to log in and use the Portal product to check out and if necessary change the Portal metadata repository values.
1. Go to the Portal "Builder" page and select the Administer Tab
2. Select the "Global Settings" link within the Services portlet.
3. Select the "Cache" tab
Review the settings for the WebCache Invalidation Host and Port.
AS10g 10.1.4:
In 10.1.4.x this page will only show you the settings and will no longer allow updates.
The alternate place to view and update these values, for all 10.1.4.x versions of Portal, is from the local Application Server Control (ASC).
1. Go to the "portal:portal" component
2. Select "Portal Web Cache Settings"
Review and make any changes here.
FMW 11g:
Changes to the Portal Metadata can be implemented using either Fusion Middleware Control (FMC) or the WebLogic Scripting Tool (WLST).
Example: Using FMC:
- Ensure the Portal Manager Server is started or at least one member of the Portal Cluster is running.
- Expand the Farm --> Portal --> portal (WLS_Portal) --> right click --> Portal wire configuration.
Example: Using WLST:
$ cd $ORACLE_HOME/common/binwith ORACLE_HOME pointing to the Portal software directory
$./wlst.shWLS> connect('weblogic','manager1','t3://portal.oracle.com:7001');
WLS:/mydomain/serverConfig> setPortalWebcacheConfig(dad_name='portal1',host='foo.oracle.com',inv_port= '6523',inv_user= 'invalidator',inv_passwd=' invalidator')
or simply:
WLS:/mydomain/serverConfig> setPortalWebcacheConfig('portal1','foo.oracle.com','6523', 'invalidator',' invalidator')
Note:
These changes are stored in the central Portal metadata repository database. For configurations with multiple mid-tiers, this change only needs to be done on one of the mid-tiers to affect all the mid-tiers.
These changes are stored in the central Portal metadata repository database. For configurations with multiple mid-tiers, this change only needs to be done on one of the mid-tiers to affect all the mid-tiers.
It is also possible to reset / change the invalidator password at WebCache level as to make sure there is a consistency in the password value entered and execute one of the actions above.
To reset / change please read further the Web Cache Administrator's Guide.- "8 Setup and Configuration > Tasks for Setting Up OracleAS Web Cache -- Task 2: Modify Security Settings -- Point 1."
No comments:
Post a Comment