How to handle sessions with two severs on one machine?

All,
I am having a problem with session cookies being overwritten when I host two apps on one machine running WebLogic 8.1 The apps are http://myserver:7300/app1 and http://myserver:7400/app2, and each runs in its own server.
Users will often access both apps at once, in two browser windows. If the windows are different threads in the same process, the sessions collide. For Internet Explorer, this isn't usually a problem since clicking on the shortcut multiple times launches different processes by default. Some browsers (Firefox, etc.) won't let you have two windows under different processes. Attempts to launch a second window 'detect' the existing process and appear to spawn a new thread. When this happens there appears to be no way for the users to use both apps at once.
I know this is happening because of the way session cookies are stored in the browser process' memory. The session cookies appear to me to be 'keyed' by the host name or ip address of the server. Does anyone know of a setting in WebLogic so that this 'key' includes the port or context root? Is this even something which can be controlled on the server side?
Thanks for any help,
Brian

Not quite sure what your intent is, but if you want to avoid a clash how
about giving each application a different default session cookie name.

Similar Messages

How to handle security with Auto-Versions and Time Machine?

I've upgraded to OS X Lion, which of course now has the new automatic "versions" system.
Today, I created a text file in TextEdit, containing highly sensitive data that I must keep absolutely secure (well, you know -- as best I can.) The file was eventually copied to an encrypted sparse disk image (.dmg) and the original on my desktop deleted trashed and emptied.
But before that happened, Lion had already saved two copies of the file through the new versions sytem --and-- one copy in regular old Time Machine! So much for keeping the file secure. (OK -- not "copies" but delta data. Whatever. )
EDIT: OK ... I found how to "Delete all backups ..." from within Time Machine. (I feel silly for not finding that before!) But the real question here relates to manual and auto-save versions from the likes of TextEdit and any residual data that may leave lying around, despite the file seemingly having been deleted.
My questions should now be fairly obvious ...
Does "Delete all backups" from within TimeMachine also delete all data from the "versions" system? One would assume so. But I don't like to assume.
Can one somehow pre-emptively flag a file to not be recorded by versions or Time Machine?
Someone in some post noted that a file could be "locked" to prevent these issues altogether. But I can find no informtaion about such a thing, either by further googling or poking around in menus locally.
EDIT: Good grief! You spend ages trying to find something, then you post on here and it magically appears. You can click the title bar of TextEdit to produce a pull-down (that's new!), which shows an option to "Lock".
So this one is just about answered by myself -- except for the part about whether "version delta" info is also removed when "Delete all backups" is used from within Time Machine.
Any help will be much appreciated. Thanks.
Bryan.

If you're dealing with highly sensitive data, you should enable FileVault on your primary and backup volumes. That's the only way to ensure that all copies of your files will always be encrypted. Needless to say, you have to be very sure you won't lose the password. Optionally, you can store an alternate decryption key on Apple's servers, but then you have to trust Apple.
That said, in my opinion it's too soon after the release of Lion to trust FileVault completely.

Sharing Addressbook with two accounts on one machine

Hi All,
somehow I cannot find out how to share my addessbook entries with my girlfriends account on the same mac.
I already did the same with iphoto where I created an alias of the iphoto data folder and which replaced her iphoto data folder. Now she has full access rights and can also edit the pictures.
I would like to the same with addressbook but when replacing the folder "User/Library/Application Support/Adress Book" address book does not start anymore.
Can someone please give me a hint!
Many thanks and best regards
Daniel

Does this article contain anything useful?
(24465)

File Vault (FV) acces to partitions with two OS on one machine

I am still on 10.6 but just ordered a new MBP, so I can not check myself at present.
For technical reasons I would like to have 3 partitions on the MBP, two of which will be a Mac OS 10.8 systems. The HD has thus 3 partitions. lets call them OS1 and OS2 and D3.
If I turn FV on for OS1 seperately in the control pannel of OS1 and OS2 again in the control pannel of OS2, will both be able to read D3? I assume yes. correct?
Now if I turn FV on for OS1 and D3 in the control pannel of OS1 FV. I assume that when I boot from OS 2 I will not be able to access D3? or?
If this is indeed the case, is there a way around it? Like would it work to have both the users on OS1 and OS2 be identical? Or?
I look forward to your thoughts.
Message was edited by: ChangeAgent

Thanks.
Not sure I understand your answer, let me recap to see if I understood.
So any internal partition that is on the MBP will only mount if I tell it to, it will not mount automatically when the machine has started up? At that point I enter the password for the disk.

How to maintain Session between two ear files?

HI Guys
I have deployed two ear files in oc4j. One is used to open applet and other for business operations.
My problem is after calling applet my session gets invalidate. How to save session between two context
I am new in this field please help!!
the sooner the better
Rahul

Can I do this if I integrate oc4j with apache!!
Can I maintain session
Plz. Help
Rahul

How to handle dbms_xmldom with no data values.(no_data_found error in dom)

hi,
i have below block,
DECLARE
doc dbms_xmldom.DOMDocument;
node dbms_xmldom.DOMNode;
elem dbms_xmldom.DOMElement;
cur_node dbms_xmldom.DOMNode;
root_elem_data dbms_xmldom.DOMElement;
root_elem_tab dbms_xmldom.DOMElement;
root_node_data dbms_xmldom.DOMNode;
mode_elmn dbms_xmldom.DOMElement;
mode_node dbms_xmldom.DOMNode;
mode_text dbms_xmldom.DOMText;
doc1 DBMS_XMLDOM.DOMDOCUMENT;
root_node_data1 DBMS_XMLDOM.DOMNODE;
child_document DBMS_XMLDOM.DOMDOCUMENT;
child_rootnode DBMS_XMLDOM.DOMNODE;
V_CLOB CLOB;
v_doc CLOB;
v_EMP CLOB;
v_output_filename VARCHAR2(300) := 'SPOOL_DIR/'||'EMP_XML_FILE.xml';
l_xmltype XMLTYPE;
BEGIN
doc := dbms_xmldom.newDOMDocument;
node := dbms_xmldom.makeNode(doc);
dbms_xmldom.setversion(doc, '1.0');
dbms_xmldom.setCharset(doc, 'UTF8');
elem := dbms_xmldom.createElement(doc, 'PartnerInfo');
dbms_xmldom.setAttribute(elem,'xmlns','EMP');
cur_node := dbms_xmldom.appendChild(node, dbms_xmldom.makeNode(elem));
mode_elmn := dbms_xmldom.createElement(doc, 'EMPLOYEE');
mode_node := dbms_xmldom.appendChild(cur_node,dbms_xmldom.makeNode(mode_elmn));
BEGIN
SELECT value(e) INTO l_xmltype
FROM TABLE(XMLSequence(Cursor(SELECT * FROM EMP1 where EMPNO=7501))) e;
child_document := DBMS_XMLDOM.newDOMDocument(l_xmltype);
root_node_data1 := dbms_xmldom.importNode(doc,dbms_xmldom.makeNode(dbms_xmldom.getDocumentElement(child_document)),TRUE);
root_node_data1 := DBMS_XMLDOM.appendChild(root_node_data, root_node_data1);
EXCEPTION
WHEN OTHERS THEN
Dbms_Output.Put_Line('Error in SELECT stmt(UC_PARTNER_MS):::'||'error::'||SQLERRM);
END;
dbms_lob.createtemporary(v_doc, true);
dbms_xmldom.writeToClob(doc,v_doc,'UTF8');
v_EMP:= v_doc;
dbms_xmldom.writeToFile(DOC,v_output_filename,'UTF8');
dbms_xmldom.freeDocument(doc);
--Dbms_Output.Put_Line('THE OUTPUT IS::'||V_EMP);
EXCEPTION
WHEN OTHERS THEN
Dbms_Output.Put_Line('Error in SELECT stmt(UC_PARTNER_MS):::'||'error::'||SQLERRM);
END;
The xml file is 'EMP_XML_FILE.xml'
<empno>U++kYmcVuGchxbh+++++++++++++++1+</empno>
<empname>J</empname>
suppose the empno 7501 is not available in our emp table,
i got error
ORA-03113: end-of-file on communication channel
how to handle xmldom with no data values.
by
siva

hi,
please give the solution
by
siva

Hi my name Abukar I had an old apple ID and I had problem with signing it so I decided to make a new apple ID with a new email address, so how can I link with two apple IDs and how can I get back all my previous apps that I purchased before, I cloud stuff

Hi my name Abukar I had an old apple ID and I had problem with signing it so I decided to make a new apple ID with a new email address, so how can I link with two apple IDs and how can I get back all my previous apps that I purchased before, I cloud stuff

It is not possible to do that.
Allan

Firefox start with two pages. One says welcome to Firefox etc. And the other is my homepage (google). I want to stop the first page. How can I do this??

Firofx (5.0) starts with two pages. One is the welcome to Firofox and also that there are Add-ons available. The other one is my homepage (google). I want to stop the first page. How can I do this?

See these articles for some suggestions:
* https://support.mozilla.com/kb/Firefox+has+just+updated+tab+shows+each+time+you+start+Firefox
* https://support.mozilla.com/kb/How+to+set+the+home+page - Firefox supports multiple home pages separated by '|' symbols
* http://kb.mozillazine.org/Preferences_not_saved

How to Persis session between two ear files

Can anybody tell me how to persists session between two ear files?
Please Help its urgent!!
Rahul

Rahul -- there's nothing as a J2EE spec requirement or even a proprietary feature of OC4J/OracleAS which will enable this.
If you have two different EAR files, then they are two different applications, and each will have their own session which can't be shared amongst each other.
If you have data which needs to be shared across different applications then you could either put it on a shared file system or write it to a database which you can then read from in the other application.
One way to do this is to implement the HttpSessionAttributeListener interface which will be called when object are added/removed/updated from a session, and then make a determination on whether the new object needs to be stored in a common place to facilitate sharing between applications.
This is more likely to be a hackaround and ideally you'd examine why you want to share session state across applications and rearchitect so you don't need to do it -- but if you want to try a solution without a major revisiting of the whole application, then perhaps it's an option.
cheers
-steve-
cheers
-steve-

How do I set up two libraries on one pc so that we have two separate libraries for each iPhone?

How do I set up two libraries on one pc so that we have two separate libraries for each iPhone?

Launch it with the Shift key held down, or create a second Windows user account.
(60390)

How to handle tree view control in business one ui sdk

Hi,
Can any guide me on how to handle tree view control in business one ui sdk?
Thanking in advance.
With Regards,
Ram.

Hi Ram,
Nowadays there are some trouble with the treeview controls in Windows XP SP2 as you can see here:
It is said that SAP will publish a treeview control in 2005 SP1.
Hope helps,
Ibai Peñ

How to install oracle817 and form6i on the one machine

How to install oracle817 and form6i on the one machine?
Machine is P4/W2K
Thanks a lot

You can do this as long as you install them in different Oracle homes.

Two FQDN on one machine.

I am trying to consolidate services on to as few machines as possible. Since you can setup multiple IP addresses, is it possible to have two FQDNs on one machine? I want to use a server to host files on my private network but also provide a public website. I already have the entries in DNS. I just don’t know if I can have two FQDNs on the same machine. It is as simple as just adding the new address?

If you are just talking about web hosting for two fqdn, then add another site in ServerAdmin:Web
Apache suppport virtual hosts. You could have hundreds of FQDN running under a single IP address.
In your case, you don't even need two IPs, but you could use them- or at least use them transitionally until you modify DNS and then just use on public IP.
If you want to add a second IP to the machine, you duplicate the port in SystemPrefs:Network. Show Network Port Configurations, then duplicate the ethernet interface and give that interface it's new settings.
Jeff

How to manage itunes with two ipods

I have two iPods, a 60G photo and a 30G video. I have two macs, one at work, the other at home. My iTunes on the work computer has about 50G of music, and about 15G on my home computer. Until now, I have only synced the big iPod with the work computer and small iPod at home. I just got a very big external drive, and want to keep everything there. That raises several questions I thought someone here might know how to answer.
1) Where are the files so I can just copy them onto the ext hard drive? (I can probably find this for myself, but with my computer skills on mac, it will be frustrating, I imagine. Since I have the other questions, I thought htis one might be easy.)
2) How can I merge the two files? I know I have some duplicates, adn I know how to get rid of them once the two are merged, but how do the two files merge together in one iTunes file on my new hard drive?
3) Once the two are merged, can I set it so that certain types of music are put on one iPod, and other types on the other?
Thanks a lot, in advance.
Brad

1. Click on the folder in the Finder's sidebar with the house icon, and then open the folder named Music. The iTunes folder inside is the default location of the music library; if the songs aren't here, follow these instructions to locate them.
2. Drag the files into the Library section of the open iTunes application.
3. Use the option in this article which best fits your needs.
(22256)

Some basic questions how to handle Exceptions with good style

Ok I have two really basic Exception questions at once:
1. Until now I always threw all exceptions back all the way to my main method and it always worked well for me but now I am sitting at a big project where this method would explode the programm with throwings of very many exceptions in very many methods. So I would like to know if there is a more elegant solution.
2. What do I do with exceptions that will never occur?
Lets say I have a method like this:
void main() {calculate();}
void calculate()
sum(3,5);
void sum(int a,int b)
MathematicsA.initialise("SUM"); // throws AlgorithmNotFoundException, will never occur but has to be handled
}So what is the most elegant way?:
h4. 1. Ignore because it will not happen
void sum(int a,int b)
try {MathematicsA.initialise("SUM");}
catch(AlgorithmNotFoundException) {}
}h4. 2. Print stacktrace and exit
void sum(int a,int b)
try {MathematicsA.initialise("SUM");}
catch(AlgorithmNotFoundException e)
e.printStackTrace();
System.exit();
}h4. 3. throw it everywhere
void main() throws AlgorithmNotFoundException, throws ThousandsOfOtherExceptions {}
void calculate() throws AlgorithmNotFoundException, throws HundretsOfOtherExceptions
sum(3,5);
void sum(int a,int b) throws AlgorithmNotFoundException
MathematicsA.initialise("SUM");
}h4. 4. Create special exceptions for every stage
void main() throws MainParametersWrongException
try {calculate();}
catch(Exception e)
throw new MainParametersWrongException();
void calculate() throws SumInternalErrorException
try {sum(3,5);}
catch (SumInternalErrorException)    {throw new CalculateException();}
void sum(int a,int b) throws SumInternalErrorException
try
MathematicsA.initialise("SUM");
} catch (AlgorithmNotFoundException e) {
throw new SumInternalErrorException();
}P.S.: Another problem for me is when a method called in a constructor causes an Exception. I don't want to do try/catch everytime I instantiate an object..
Example:
public class MySummation()
     Mathematics mathematics;
     public MySummation()
             mathematics.initialise("SUM"); // throws AlgorithmNotFoundException
     void sum(int x,int y)
          return mathematics.doIt(x,y);
}(sorry for editing all the time, I was not really sure what I really wanted to ask until i realised that I had in fact 2 questions at the same time, and still it is hard to explain what I really want to say with it but now the post is relatively final and I will only add small things)
Edited by: kirdie on Jul 7, 2008 2:21 AM
Edited by: kirdie on Jul 7, 2008 2:25 AM
Edited by: kirdie on Jul 7, 2008 2:33 AM
Edited by: kirdie on Jul 7, 2008 2:34 AM

sphinks wrote:
I`m not a guru, but give my point of view. First of all, the first way is rude. You shouldn`t use try with empty catch. "rude" isn't the word I'd use to describe it. Think about what happens if an exception is thrown. How will you know? Your app fails, and you have NO indication as to why. "stupid" or "suicidal" are better descriptions.
Then for the second way, I`ll reccomend for you use not printStackTrace(); , but use special method for it:
public void outputError (Exception e) {
e.printStackTrace();
}It`ll be better just because if in future you`d like to output error message instead of stack of print it on GUI, you`ll need change only one method, but not all 'try' 'catch' statements for changing e.printStackTrace(); with the call of new output method.I disagree with this. Far be it from me to argue against the DRY principle, but I wouldn't code it this way.
I would not recommend exiting from a catch block like that. It's not a strategy for recovery. Just throw the exception.
Then, the third way also good, but I suppose you should use throws only if the caller method should know that called method have a problem (for example, when you read parametrs from gui, so if params are inccorect, main method of programm should know about it and for example show alert window)."throw it everywhere"? No, throw it until you get to an appropriate handler.
If you're writing layered applications - and you should be - you want to identify the layer that will handle the exception. For example, if I'm writing a web app with a view layer, a service layer, and a persistence layer, and there's a problem in the persistence tier, I'd let the service layer know what the persistence exception was. The service layer might translate that into some other exception with some business meaning. The view layer would translate the business exception into something that would be actionable by the user.
And in all other cases I suppose you should use the fourth way. So I suppose it`s better to combine 2,3,4 ways as I have mentioned above.I don't know that I'd recommend special exceptions for every layer. I'd reuse existing classes where I could (e.g., IllegalArgumentException)
Sorry, I can give you advice how to avoid problem with try-catch in constructor. :-(You shouldn't catch anything that you can't handle. If an exception occurs in a constructor that will compromise the object you're trying to create, by all means just throw the exception. No try/catch needed.
You sound like you're looking for a single cookie cutter approach. I'd recommend thinking instead.
%

How to handle sessions with two severs on one machine?

Similar Messages

Maybe you are looking for