I have an RP for which I've built a login page using the Json feed from ACS. The IP images are linked to the .LoginUrl attribute of the feed and when I click on one of the images it correctly jumps to that IP's page.
Entering my credentials, however, I'm redirected to a page on the appfabriclabs.com site with the following error:
HTTP Error Code: 400
Message: ACS50000: There was an error issuing a token.
ACS50011: The RP ReplyTo address is missing. Either the RP ReplyToAddresses
are not configured or an invalid wreply 'https://www.skillscore.it/' was received
in the sign-in request.
the RP is configured in the App Labs site with a returnUrl of:
https://www.skillscore.it/Home/FederationResult
and in looking at the wreply parameter in the feed, I see:
https%3a%2f%2fskillscore.accesscontrol.appfabriclabs.com%3a443%2fv2%2fwsfederation
According to some SO articles like [this one] the return url of the app should be a prefix of the wreply parameter - which is clearly not the case here.
so... what have I done wrong now?
e
p.s. one interesting bit of info: in the Application Integration page of ACS there is a link to the ACS-hosted login page. the link used there seems to differ from the one I'm given in the feed; in particular, the ACS-hosted page uses a wctx of:
pr%3dwsfederation%26rm%3dhttps%253a%252f%252fwww.skillscore.it%252f
whereas the feed gives me:
pr%3dwsfederation%26rm%3dhttps%253a%252f%252fwww.skillscore.it%252f%26ry%3dhttps%253a%252f%252fwww.skillscore.it%252f
so I don't know what that's worth but maybe it's a clue to what's wrong.
* update *
decoded, that last string is:
pr=wsfederation
&rm=https%3a%2f%2fwww.skillscore.it%2f
&ry=https%3a%2f%2fwww.skillscore.it%2f
which clearly shows the Json feed is providing an ry that is not present in the ACS-hosted page... meaning anything to anyone?
ok. my bad. apparently, when I was fetching the Json feed, the URL I used did not have the reply_to set correctly.
Related
I have tried below but http request got fails is there I missing something can anyone guide me please. In postman by using qnaId in the body I am getting the correspond queastion and answer in json response. The same I have tried in composer with Http request but it fails.
{
"qnaId" : "1164"
}
enter image description here
enter image description here
The same I tried in Postman and successfully got the response.
enter image description here
Thanks....
I think the issue is that $Qna is not a valid memory variable. Per documentation, you need to set this in either conversation, dialog, turn or user scopes. I had missed the section in the post here where the response was being set to $Qna. That's not valid. So, it needs something like turn.Qna or dialog.Qna.
Additionally, the use of $ should be used only for string interpolation, not in setting the variable name. turn.Qna would set it and you would utilize it in a string as: "You said: ${turn.Qna}"
I have to do load testing for a web based application. I am getting status as warning the results table. My request contains the URL and the path as /. I have passed username and password in the parameters section. Even after tried many times still it shows status as warning.
I have also tried using Proxyserver address, port, user name and password ...still no luck..
Please help me on this.
If you get a Warning status, this means that JMeter detected a response code > 399.
There can be a lot of reasons for this, examples:
Wrong URL : 404
Error : 500
To have more details on it, add a View Results Tree and inspect all tabs to see:
Request : What you are sending (headers / Cookies / body)
Response : What you are getting (headers / Cookies / body)
Then fix your HTTP request by comparing request in browser with what you have build.
Alternatively, use JMeter recording feature.
To see all ways to debug a script, have a look at this book where sample chapter explains lot of ways.
If you get warning status that means your entered URL contains https:// part which means your given URL working with HTTP protocol.
Remove https:// part from
- HTTP Request Defaults
- HTTP Request
So replace https:// part by WWW and try again
I'm using Bottle microframework (but I doubt my issue comes from it)
First if I define a simple redirect from /test_redirect to /x it works. So Bottle redirect() on simple case works.
Now I have a page /buy that uses Stripe Checkout (custom form) pointing to my server /stripe_process
Basically Stripe Checkout verifies credit card transaction, creates a token and POST it to my /stripe_process.
/stripe_process does its stuff (calling Stripe to charge the card) then when transaction is success, i use instruct Bottle to:
redirect('/transaction_summary')
My webserver logs show that indeed /transaction_summary is called and server-side script is processed (I put flags in my script to check this), and its template is returned (browser 'Network' analysis gives a 303 then a 200 on /transaction_summary, I even get the correct response data: seems 100% normal), but on the browser page nothing happens (I still have my initial page /buy in url field, instead of having been redirected to /transaction_summary): browser received the /transaction_summary response, a preview is visible on Chrome devtools preview mode, but is not displaying it!
Also if in /stripe_process I change redirect() with a simple return template(), same issue: no data is processed by the browser.
Any clue ? This behaviour have been observed on 100% my tests:
- Firefox / Windows
- Chrome / iOS 9 (iPad)
- Chrome / linux
- Iceweasel / linux
- Chrome / Android
- Chrome / OSX
I suspect it has something to do with Stripe Checkout taking over something (since redirect() works perfectly in my simple test), but I can't figure the reason and how to solve this.
If from Chrome-devtools-network section I go to last action (ie /transaction_summary download) and open /transaction_summary in new tab, it renders perfectly.
So it's not a bottle problem, nor webserver. I suspect more on Stripe checkout modal/popup behaviour
$(window).on('popstate', function() { handler.close(); }); is present as stripe docs instruct but anyway stripe_handler is properly closed after token received (I checked with handler.closed callback...)
EDIT:
If I replace redirect() with straight return template(), same issue: html stuff is downloaded by the browser, avail on preview, but not rendered on main window.
EDIT2:
if I add on my page a href to /test_redirect redirecting to /x it works.
Note that this manual redirect work after Stripe checkout. The only difference with Stripe sequence here is user interaction ('click' on href) but as my redirect is same domain, browsers shouldn't block redirect anyway.
Problem was that AJAX call used to send Stripe token to /stripe_process was handling redirect response. That's why on network I had 200 answer from webserver but all html page data was going into the AJAX callback and rendered. Thanks to Thomas for raising my nose from this issue.
Solution is /stripe_process to return 1 in case of success (instead of server-side redireciton), then make the redirection from AJAX success callback.
I fire an image tag when someone converts on our website for reporting and analytics:
<img src="https://example.evyy.net/conv/?somechannel=cats&cid=3790&oid=123&cat1=premium&sku1=123&qty1=1&amt1=456&custid=qbc" />
We noticed traffic decline yesterday. To debug I visited the site and signed up, with the console open network tab.
I typed "evyy" into the filter box and I see 3 line items for my tag.
What does this mean? Should there not be just 1 line item? One item has a 302 status and the other two a status of 200. That's good right?
Screen attached. Does this look "right"? Is there any drill down in the network tab I can look at to see if there is a problem on our end with tracking?
"Initiator Other" usually means the network request was made from a user, not from Chrome, a redirect or a script.
The request was fired when someone clicked a button (this is what your analytics is probably tracking). You can read more about this in the Chrome DevTools docs.
Initiator: The object or process that initiated the request. It can have one of the following values:
Parser - Chrome's HTML parser initiated the request.
Redirect - A HTTP redirect initiated the request.
Script - A script initiated the request.
Other - Some other process or action initiated the request, such as the user navigating to a page via a link, or by entering a URL in the address bar.
The requests are made in the order you see, the first received the 302 response which said, "hey, go to this new url". That's (probably) why the second request was made, which got the 200. The third was probably from clicking on the button too. Looks fine to me.
I noticed another case not related with a user action:
<link rel="icon" type="image/jpeg" href="image.jpg" />
... causes the image.jpg to get loaded with "Initiator Other" (also true with images of another format).
(wanted to add this as a comment under the previous answer, but I am not allowed to)
One case where initiator type is 'other' is when a request gets redirected, in this scenario the redirected request will have initiator as other.
For Example: request was made for url1 which got redirected to url2, now in case of url2 we get initiator type as 'other'
I'm trying to test java web application (jsf) that uses oracle adf-faces. When I record my test, the application is working fine but when I later check the user, then on login.jsf page, just before authentication and home pages, I get 500 Internal server error that says:
ADF_FACES-30200:For more information, please see the server's error log for an entry beginning with: The UIViewRoot is null. Fatal exception during PhaseId: RESTORE_VIEW 1.
Some more info: My login.jsf page is referred by previous login page that has jsession id appended to it. However, as far as I understand, neoload automatically handles jsessionid (I did try to set a parameter for it though and it didn't fix the error)
Also login.jsf page itself is a referrer to authentication page that in its turn is a referrer to a home page of an application, and the latter are both redirect pages (do not use manual definition) so they are not played during the check because of this error:
login.jsf page uses POST method and it has the following POST parameters: pt1:username, pt1:password, org.apache.myfaces.trinidad.faces.FORM, Adf-Window-Id, javax.faces.ViewState, Adf-Page-Id, oracle.adf.view.rich.DELTAS, event, event.pt1:loginCommandButton
I use set of users (I have a variable for username and password) and I handle javax.faces.ViewState manually with a jsf parameter (it used to give me errors).
I would appreciate any help greatly!