Css stickiness based on the http header

there is CSS 11503 that should load balances the traffic between 2 servers running IIS (http port 80). In front of load balancer there is a reverse proxy, that hides all real ip addresses of users that send requests to web-servers.
The customer would like to have stickiness per user. The reverse proxy can add user real ip address to http header.
What kind of load balancing mechanism is better to use to fulfill customer requirements? HTTP load balancing? If "yes", are there standard field-types that is possible to use?

You can't do sticky on http header in CSS , best solution is to insert a cookie for stickyness, individula clients will get a cookie and will stick based on cookie ppresented.
see:
http://www.cisco.com/en/US/docs/app_ntwk_services/data_center_app_services/css11500series/v8.20/configuration/content_lb/guide/Sticky.html#wp1109390
content testing
vip address 192.168.128.131
add service s1
advanced-balance arrowpoint-cookie
active

Similar Messages

Feature Request : provide a way to create access policies or identities with matching condition based on the HTTP header's "Referer" field

Hello,
I have a use-case I would like to share with you. When a customer configures its WSA with highly restrictive internet access like in the example below, it may trigger some issues :
1- allow internet access only for URLs defined in whitelist.
2- block ALL other requests.
Let's take the following example :
1- the customer only allow requests to www.siteA.com. siteA.com is the only URL included in its whitelist.
2- www.siteA.com contains many embedded objects (such as facebook like tags, youtube videos, links to partners sites, ...)
In this configuration, the end user will be allowed to reach siteA but the page will not be fully displayed. All the embedded objects not directly located on siteA will be missing.
With WSA, the easiest way I can imagine to solve the issue is to list all the embedded objects present on siteA, get back their URL and also add these URLs to the whitelist. But this solution if of course far to be really convenient since it involves to know exactly how each HTTP page you want to consult is built.
With other proxies, such as Bluecoat proxies or McAfee Web Gateway proxies for example, I used to solve this kind of issue by using the HTTP referer field (the URL you come from). For example with Bluecoat :
<Proxy>
ALLOW request.header.Referer.url.domain=//www.siteA.com/
=> All requested objects from siteA.com will be automatically allowed by the proxy, even if they are not part of my whitelist.
- Do you have a better suggestion than the one I'm currently using with WSA (adding each sites in whitelist) ?
- Would it be possible to add the field HTTP referer as a matching condition for Identities and access policies in your next release ?
Thanks in advance
Best regards

As far as I'm aware this functionality is still not available... would be an awesome feature to have, but could also be abused at the same time by a user writing their own "middleware" proxy and setting the referrer header to that allowed site.. could be done in like ~15 lines of perl / python.
Either way... would still be a cool feature to have.

What happens to the HTTP header parameters I put there?

Our tool puts some parameters into the HTTP header. But, it appears that the Dispatcher is not passing them to the WAS. Could this be?
Is there a way to tell the Dispatcher to not drop our HTTP header stuff? Perhaps a configuration setting? I didn't find any.
When I set a break in my servlet running on NW WAS, I see:
referer:
Cookie:
etc.
but I don't see the parameters we put there.
This is the NW SneakPreview download.

Hi Farokh,
I think there should be much problem passing parameters through the Url. I am sorry however that I have not worked much on the Servlet but on WebDynpro, which I can share with you if its of any use to you.
Regards
Noufal

Webdynpro - How to retrive the http header

Hi,
Normally you will be able to retrieve the http header from the java dynpro HttpServletRequest but I have do you retrieve it from a <b>JAVA webdynpro component</b>. Have not been able to figure it out so please any help appreciated.

Hi Rasmus,
try the following to catcth the request object.
WDProtocolAdapter.getProtocolAdapter().getRequestObject()
Regards,
Shubhadip

How can i get the HTTP header info using Java

Hi,
I kinda need to extract the HTTP header info from a page that i goto through a URL objecrt and socket. Java.
Can someone please help me.. thanks!!!!
System.out.println("Enter URL : ");
String getAdd; // The URL that the crawler begins at.
// Manually take in URL
Scanner sc = new Scanner(System.in);
getAdd = sc.nextLine();
URL myURLobj = new URL(getAdd);
String hostName = myURLobj.getHost();
Socket s = new Socket(hostName, 80);

URL url = new URL(getAdd);HttpURLConnection conn = (HttpURLConnection)url.openConnection();
Map<String, List<String>> headerFields = conn.getHeaderFields();

Load runner: in which the HTTP header is stored?

I've used Mercury Load runner to stress test our Web application.
We used a session variable in the url, so after I scripted, I can't replay it (error: session is expired). So, I want to know, in Mercury Load runner, where the HTTP header is stored?

Hi,
The issue is not where the http header is stored. For make it works you just have to create a correlation to get this session id dynamically in your script.
For know how to do it have a look at the web_reg_save_param function documentation or ask google.
Regards,
Sylvain

CSS Bypassing farm traffic based on matching HTTP header

Hi,
I am trying to find out whether the CSS is able to bypass specific traffic.
I have an existing content to match all HTTP and send to a farm. However, there are some HTTP flows i dont want to goto the farm, i just want CSS to route them onward to the destination. These specific HTTP packets are differentiated by the host field in the header. What config is needed to allow these host annotated packets to bypass the serverfarm?
Thanks
Alan

Hi Gilles,
Thanks for your response. The only thing you may have misread is that i need to select the host header field, as the URL's may not have host part in them, ie. raw http, not proxied. I guess then i need a header match rule linked to the new content, instead of the URL filter you mentioned.
BR
Alan

Load balance traffic to a service based on a added field in the HTTP Header

I am trying to use HTTP Header Load balancing but the field we want to use in order to load balance is "user-defined", example HTTP_TOTO = toto1.
Do you have any idea on how I could perform this ?
Thanks in advance

Load balancing using pre-defined headers is supported. Not sure if load balancing using user defined fields is possible. You could refer to the following document.
http://www.cisco.com/univercd/cc/td/doc/product/webscale/css/css_710/bsccfggd/httphead.htm
We would appreciate it if someone could share their experience if they know more about this.

How do I set the Http Header POST URL in SAAJ?

Hi ,
I am a newbie in the field of web services. I was trying to create a SOAP request with Http Header POST information having a POST url like
the following:
POST /OMASTI.xml HTTP/1.1
Content-Type: multipart/related; boundary="eladeladeladeladeladeladeladeladelad"; type=text/xml; start=11814460
Content-Length: 54596
Host: unspecified
SOAPAction: ""
to that effect I did the following in the SAAJ client:
MimeHeaders md = message.getMimeHeaders();
md.setHeader("SOAPAction" ,"\"\"");
md.setHeader("POST" ,"/OMASTI.xml");
However what I got was :
POST / HTTP/1.1
Accept: text/xml, text/html, image/gif, image/jpeg, *; q=.2, */*; q=.2
SOAPAction: ""
POST: /OMASTI.xml
Content-Type: multipart/related; type="text/xml"; boundary="----=_Part_2_32124414.1153146262750"
Content-Length: 3352
Cache-Control: no-cache
Pragma: no-cache
User-Agent: Java/1.5.0_07
Host: www.google.com
Connection: keep-alive
I have two POSTs in the header. How do I fix it? Please Help ASAP.
Regards,
Zeus

Hi,
Please forgive my ignorance. I did the same as you said. and then checked the Http request being sent out using ethreal software... it was the same as I had told earlier. Is only POST supported by the connection object? And the does the URL that comes in the Mime header as POST refer to the URL to which the request was sent? what I mean is that if the mime header says
POST /OMASTI.xml HTTP 1.1, does it mean that the URL that was passed to the connection object was "/OMASTI.xml" ? I had given the url as http://www.google.com to test the request being sent.. I only have a sample SOAP request which shows a HTTP header with the post url as I said before. I need to create a SOAP request with the same sort of Http header and the body needs to follow a certain OMA-STI protocol. My experience in the SOAP domain is almost nil. So please enlighten me.
Thanks in advance,
zeus

How to specify more than 6 parameters in the URL or in the HTTP header?

After applying SAP note 1105368 ,it will be availible to specify more than 6 parameters in the URL header.
But i haven't figured out how to specify it yet.even though i went through the note carefully, but it really hard to understand.
Is there someone who experienced to specify more than 6 parameters ?
Inccording to the note,the parameter fileds look like should be
specified like below.
parameter1: <SAP:Record namespace="http://sap.com/xi/XI/System/HTTP" name="SAP_URL_EXTENSION">urlext</SAP:Record>
parameter2:<SAP:Record namespace="http://sap.com/xi/XI/System/HTTP" name="urlext01">Order=4711</SAP:Record>
parameter3:<SAP:Record namespace="http://sap.com/xi/XI/System/HTTP" name="urlext02">Object=1807</SAP:Record>
but something about the format doesn't look right....

Hi,
I know we can set headers by using some tools. Neoload tool is used to set values into header.
If Servlet after receiving the request you can set header values and send back with response.
Check this
http://www.unix.com.ua/orelly/java-ent/servlet/ch05_06.htm
Thanks,
RamuV

IDoc to File with different file name and data based on the IDoc header dat

Hi,
I got a requirement to implement IDOC to file senario. In this the Converted file should have the file name and file content based on IDoc Header data. Means If one of the IDoc header field value is
'A' - then the file name recieved by reciever system should be 'A.txt' and with the columns A B C D F.
'B' - then the file name recieved by reciever system should be 'B.txt' and with the columns C D F.
In both the case IDoc message type is same.
Please suggest me what i need to do to achieve this.
Thanks in Advance. ...
Regards
Sri

Hi,
As per my requirement, we need to create one file for each IDoc based on IDoc header data. SAP program generates 10 IDocs for each run. All 10 IDoc's message type is same. Each IDoc has one header record with the file name. Base on the name in the IDoc header record, file needs to be created with different columns and file name. Means..
IDoc 01- Header Description contains 'A' - than file will be generated should have name as 'A.txt' and columns 'parent' 'child' 'descripion'.
IDoc 02- Header Description contains 'B' - than file will be generated should have name as 'B.txt' and columns 'Name' 'Department' 'Description'.
IDoc 10- Header Description contains 'J' - than file will be generated should have name as 'J.txt' and columns 'Order No' 'Materail' 'Description'.
Hope this will give more clarity on my requirement. Please let me if you are still not clear on requirement.
Thanks for your all help..
Thanks & Regards
Sreeni

Javafx media component can not modify the http header. javafx media component can not append parameter in url such as m3u8?channel=2

1. When our HLS stream server check user name and password. In http header such as the cookies value. There are not place for it.
2. The mp4 can append ?parameter=value. But m3u8 can not be appended.

1. When our HLS stream server check user name and password. In http header such as the cookies value. There are not place for it.
2. The mp4 can append ?parameter=value. But m3u8 can not be appended.

I`m trying to develop an optimistic concurency based on the Etag Header. This header is being sent OK on every GET request, but NOT being sent on a POST. Why?

I`m trying to implement an optimistic concurrency control over a web application entities.
I`m attaching an Etag HTTP Header to every URI that represents an entity.
On successive GET requests this header is being sent back as If-None-Match (witch is OK)
The PROBLEM is on a POST request, the ETAG is not being sent back (neither If-None-Match, If-Match, etc..).
My question is : Is this the desired behavior? The browser omits this on purpose on a Http POST?

Hi Daniel,
Do you still have this problem ?
(If you found a solution a link to that may help others)

Can JMS Message be filtered dynamically based on the JMS Header variable

Hi All,
We have a requirement to filter messages from an MQ (Queue) using the JMS Adapter.
There is only one subscriber to this Queue.
Can we use the Message Selector option using the same subscriber to pick different messages based on the context set.
For example I have two BPEL instances running for the same BPEL process and both of them have to read messages from the same MQ Queue, say SampleQueue.
The first BPEL instance need to pick the messages when JMSpriority = 1
The second BPEL instance should pick the messages when JMSpriority = 2
Is this possible? As the Message Selector in the examples given is expecting a literal on the right side of the expression.
Ex : Country='UK' Color='Green' etc
To put the question in a different way, can the right side of the expression in a message selector be a BPEL variable?
Thanks

Hi
I have the same question. Basically, I want to match my response message with the request message. I am putting in an id in JMSCOrrelationID in the request. The applciation that send the response put the COrrelationId, I sent into the JMSMessageId. But now, I want to correlate my request and response? What is the best way to do this in BPEL?
Can the right side be a variable? I want to say JMSMessageID = 'ID: ' + JMSCorrelationId.

Can CSS route based on cookie info in HTTP request

Hi
I am new to CSS and am interested as it might be able to provide a solution to a problem I have seen.
We currently have 3 Windows Servers running an ASP-based web application with a clustered SQL Server backend. The front end uses windows load balancing to distribute the load. All 3 servers are configured the same and there is only one application.
The problem lies with the way an the application maintains session state. I am told it uses a non-persistant cookie on the client which corresponds to a session object on the server. This is opposed to maintaining state in central location such as the database. Obviously this means the client needs to be stuck to a particular server for that session. This is currently achieved by setting the Windows NLB to single affinity which places traffic from a particular IP address to the same server. This does work but the clients source IP is changed by a downstream firewall to a NAT overload address meaning all clients appear with the same IP address (different port) and hence always end up on just one server.
The obvious next step in my mind would be to change the way NAT is done but this is not possible. The next obvious idea would be to change the application so that it maintains state in the database so the affinity of the Windows NLB could be disabled meaning requests would be dealt with using the source IP and port and hence distributed evenly. I am told this cannot be done either :) Joy!
So I have begain to look at other possible solutions. Apologies for my very limited knowledge on the CSS as I am trying to get my head around how it can be configured. I am thinking that it may be able to help me if I used it instead of Windows NLB. I am interested in the way you can use Layer 5 stickiness. Would it be able to examine the cookie in the HTTP request and route the traffic to the correct server?
I am aware that this will not alleviate the failover issue. If one of the servers were to fall over then the client would have to login again, however I am under the impression that this is acceptable behaviour. The main driver here is to provide load balancing to improve application performance by using all resources opposed to just one.
Many many thanks to anyone who can give me advise on this.

Hi Gilles
Thanks for the info. Sounds like we are on the right path. Unfortunately I am unable to get much information from the developers. Long story. I did logon to the system whilst doing a network trace. The following is what I found in the HTTP header:
Hypertext Transfer Protocol
GET /XXXXX/Includes/style.asp HTTP/1.1\r\n
Accept: */*\r\n
Referer:
http://xxx.xxxxx.xxx/xxxxx/login.asp\r\n
Accept-Language: en-gb\r\n
Accept-Encoding: gzip, deflate\r\n
User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; .NET CLR 1.1.4322)\r\n
Host: xxxx.xxxxx.xxxxx\r\n
Connection: Keep-Alive\r\n
Cookie: ASPSESSIONIDQCBCDSCR=AEHBCJEDDGMMCCBHBICLELGD\r\n
\r\n
Why do you configure 2 services? How would I go about this given the cookie in the HTTP request?
Many thanks
Gary

Css stickiness based on the http header

Similar Messages

Maybe you are looking for