I have a QTP script that interacts with a webpage. Occasionally, this webpage will pop-up dialogs that interfere with the rest of the test, and need to dismissed with a simple "OK" before I can continue running the test.
Recovery Scenarios are supposed to take care of this exact situation, but try as I might, QTP will never recognize the pop-up. The recovery scenario is set to trigger upon seeing a dialog with "Window title: Message from webpage". I used the object spy so I know that this is the correct identification.
Attached is the dialog box that the webpage occasionally displays. Very simple dialog, so why won't QTP recognize it?
Turns out, my Recovery Scenario was built correctly, but the problem I had was that QTP didn't know when to activate the Recovery Scenarios.
Go to File > Settings > Recovery, highlight the recovery scenario you want to have activate, then change Activate recovery scenarios to On every step.
By default, the recovery scenarios are only activated when QTP throws an error, not when your web page throws an error.
Lastly, having the recovery scenario check every step will probably degrade performance.
You should inspect when this window appears like a page refresh, opening a new window or clicking on a button, or operating on a page for 2 long.
Then since this is a standard window you get, why don't you add in Object repository.
Then create a function which checks for this window's existence
call this function when any of the suspected scenarios happen.
this will not degrade the performance badly comapred to recovery scenario.
Related
If I wanted to check if an application is already open or not, how can we check that with process exists action in utility VBO in Blueprism? It is asking for process name when I go with process exists action as input. I have given a process name where we can see it in the task manager but it is not producing the desired results.
For an example, if I want to launch a Google page through BluePrism but it is already open, now I want to know whether this Google page is open or not. For this we can use Utility VBO but I am not getting it.
How to fix this problem?
"Process exist" action allows you to check if the process with this name is running right now in windows. To use that, you need to have a proper, process name. For example for internet explorer it's "iexplore", and for MS Excel it's just "excel".
Using this action you can get only understanding if the process exists or not, and nothing more.
If you'd like to know if the webpage google.pl had opened, then you'll need to use other techniques - namely using application modeler, spying windows and using a variety of blueprism actions.
While launching any web application, BOT should attach the process to window/application, if attach action returns false then only BOT need to launch the site in IE. For Attach action there will be different inputs are available some of them are listed below:
ProcessName
WindowText
Trying to spy SAP system through BluePrism for SAP Automation.
When I click Identify button from BluePrism application modeller and move my mouse over to the T-Code textbox , then the text Box is highlighted using a red color box which says the TCode element is recognize by BluePrism tool in SAP test system. however the same is NOT identified for SAP Production system.
We checked both SAP system has same settings maintained to enable SAP GUI Scripting.
Appreciate help!
I don't know if it solves your specific problem, nevertheless:
Pressing alt multiple times will change the spy method, and you'll be able to find what you're looking for (even a method specifically for SAP).
please check if the SAP Id of the item is the same in both systems. You can do it for example by identifying item in both systems. Please return with the answer what's the difference in ID between the systems.
If you are starting out with SAP Automation in your organisation, you will probably not have GUI scripting enabled in the Production system, but it may be in the DEV system.
If you can access SAP then run RZ11 and enter sapgui/user_scripting then it will show a list of values, and the current value should be true. Compare DEV and PRD. You will probably need the SAP Basis admin to change this.
Read this blog for more details
Andy
For the best SAP automation performance with Blue Prism you need to enable both Client and Server scripting
i. Enabling Client-Side Scripting
Start the SAP Logon and log in to the SAP server.
Click the 'Customize Local Layout' toolbar button (ALT+F12).
Click Options and in the GUI Options screen select the 'Accessibility and Scripting' node
In the ‘Accessibility’ node ensure that the ‘Use Accessibility mode’ option is enabled.
In the 'Scripting' node ensure that the ‘Enable scripting’ option is the only option set. Notifications must be switched off to prevent popups from blocking the spy mode.
NOTE: If the options 'Notify when a script attaches to SAP GUI' check box and the 'Notify when a script opens a connection' are left checked then when you spy an element in the SAP interface you will see the message “A script is trying to attach to the GUI” every time you attempt to click on it. If you see the message “Scripting support is disabled by the user” then the ‘Enable scripting’ option has not been checked.
Save the settings and restart the SAP GUI. The clientside settings have now been set.
ii. Enabling Server-Side Scripting
Start the SAP Logon and log in to the SAP server.
Run the ‘RZ11’ transaction (type ‘RZ11 ‘and click the green tick button)
Type ‘sapgui/user_scripting’ in the Maintain Profile Parameters window.
Click the ‘Display’ button
Click ‘Change Value’ in the Display Profile Parameter Attributes window.
Type ‘TRUE’ in the ‘New value’ field.
Save the settings and log out from SAP Logon.
Exit from the SAP Logon program (click the red cross button in the toolbar, or click the ‘X’ window button to close the window and the application). You may be prompted to save any unsaved changes, if you haven’t done so already.
I am using GitLab in combination with Sourcetree. Every time I want to commit something, I would go to my Domino Designer and right click the database, click Team Development and use 'Sync with ODP...'.
Most of the time everything works perfectly fine but sometimes the synchronization dialog pops up and just wont go away. In my Sourcetree I can see that some files are waiting to be commited but those files are not all.
So the dialog shows up, wont go away and says "Progress information" "Exporting..." and thats it.
There is no "use in background" button there and the 'cancel' button is disabled.
At this point the only thing I can do is to shut down the Designer using the task manager but the problem still wont go away even after restarting my computer and VirtualBox which I am working on (The designer runs on the VM). The only thing that changes: It sometimes gets stuck "later" in the progress.
I have read the designer hangs often due to automatic synchronization. In my case this feature is disabled.
Is there someone who can tell me what causes this problem and how I can fix it?
- Thank you in advance
(The synchronization is in progress for roughly four hours now, with no change)
Are you able to track it down to a specific design element? If there is a corruption in the DXL, the export to the ODP will probably fail. In Package Explorer you can use the "Open With..." right-click menu option to open as XML. Similarly, using the menu options to export as DXL might also highlight any corruption in DXL. Also, is the NSF local or on a server. If it's on a remote server, that may impact the performance of syncing to an ODP (which will be local).
We are using Domino Designer and Sourcetree in combination with Mercurial and Bitbucket for about a year now.
We suffered a lot of difficulties: "hanging" sync, corrupted custom controls, missing files, lot of stress etc.
What I did find out: most of the syncing problems have something to do with the file IconNote in the resources.
Sometimes the associated source contains another version of this resource, so it creates an IconNote.orig. Whenever this file exists, the syncing will go wrong. I guess that IconNote.orig isn't allowed in Resources and so the syncing stops without any warning(!) and often corrupted files.
It took me a very long time find out, but now I know: If IconNote exist in the Git, I delete it instantly! Syncing goes well aftwerwards.
Scenario:
User preview xpage for editing in web browser.
Developer replicate a local replica to the server.
User click save button and trigger save action partial/full refresh
In 8.5.2 crash the whole database based on can't find java design classes. In 8.5.3 no error CS/SS accrued but there are no changes applied to the document. Seems like if you replicate the sessionID is overridden. Is there a way to fix/detect it?
Any ideas?
thx
It relates to this effect: Meaning of java.lang.ClassCastException: someClass incompatible with someClass.
Simply said, every design change resets XSP engine. Since 8.5.3 you can control it by property "Refresh entire application when design changes" in XPage properties - turning it off (default since that version) will just "soft reset" XSP engine.
Anyway, you may loose some scoped variables and beans. This concludes to some rules to obey:
Do not allow developers to change design in production during working hours.
Notify users about the problem (onError event for partial refresh) - Tim's comment explains, that you won't get an error from partial refresh after design change, unfortunately. Simple CSJS code pinging app availability and report problem when app does not respond could work, but I don't advise that: we use "keep alive" control from ExtLib, and troubleshooting in domlog.nsf became a nightmare - it is full of pings (almost all the time - many users keep their browsers open overnight).
If data are very important, implement auto save mechanism similar to Google apps.
BTW: This effect is similar to time-outed session - partial refreshes will fail and user need to reload the page (and possibly loose edits).
I'm experiencing problems when trying to deactivate a custom feature through the GUI. When I press the 'deactivate' button I'm, as expected, redirected to the warning page which asks if I'm sure I want to deactivate.
Upon confirming, the page starts loading.
The feature in question should normally be activated very fast, however on this occasion the page loads for more than 5 minutes without anything happening.
After concluding that the page seems to be stuck in an eternal loading cycle, I had to refresh the page to see if there had been any changes, but no, the feature remains active.
Any ideas?
Details:
The site I'm working is a previously existing office365 site. I've just made some changes to my custom solution (modifying one feature and adding another)deactivated the old solution and uploaded the new solution, so I'm trying to deactivate and reactivate the feature which I've modified.
It depends what's happening during deactivation - 5 minutes might not be very long if the deactivation has to perform a lot of updates (e.g. removing columns from lists in lots of subsites). I did once have a feature deactivation that took 45 minutes (?) to run.
I guess another possibility is that you've got C# code that contains an infinite loop? Though that seems a bit of a long shot.
Otherwise, Office 365 is very hard to debug; I'd suggest raising a support call with Microsoft through the O365 portal to see if they can see any logs.
When the feature is removed are you trying to remove items like CT's or anything that is dependant on another thing?
I've seen features get glitchy when this occurs on our test enviroment and make sure this is ironed out before deploying to live.
Theres a feature checker somewhere on codeplex but i'm not sure if this will connect to the office365 site. It's called something like FeatureAdmin.exe This might help you in removing the feature and clearing out left over stuff from the feature however it will not remove whatever your feature is struggling to remove(if it is of course!)
When the original feature was deployed did you test it could be deactivated cleanly?