I want to load test an enterprise Web application (which I did not build), using a Visual Studio 2010 Ultimate Load Test. I want each virtual user to log in at the beginning, and log out at the end of their run of random tests. I can properly configured the load test to do so. However, there is a complication. The session key is injected into the URL, like this:
http://ProductName/(S(ilv3lp2yhbqdyr2tbcj5mout))/System/Container.aspx
I converted the Visual Studio WebTests to coded tests, and then retrofit them with code that uses the session-specific URL. This works fine. What I need to do is persist this session encoded URL across the various tests that specific virtual user runs, starting with the login WebTest class, to the logout WebTest class.
The individual WebTest classes are capable of logging in and out at the beginning and end of each test. However, this is not an accurate representation of normal use. This application emulates a mainframe terminal, and never cuts the connection or session between Web browser requests. Each session is one long, interactive HTTP request, just like a mainframe terminal interacts with, for example, an IBM AS400. Usert typically log in to the mainframe at the beginning of day, and (should) log out at the end of day. Likewise, this Web application maintains the HTTP request until the user logs out, or the IIS session timeout occurs. Therefore, it is important I keep the same session in the URL, between all tests, to ensure memory leaks and other nasty bugs don't accumulating.
Please share your thoughts!
Problem 1: persist the session id across test iterations
You can store data in the 'user context' which is persistent across test iterations. It is found in the WebTestContext having the name '$LoadTestUserContext'. (But note that this context parameter only appears in load test runs, not in standalone web test runs)
// within WebTestPlugin.PreRequest() or MyExtractionRule.Extract()
// e is the corresponding eventargs object...
LoadTestUserContext userContext = (LoadTestUserContext)e.WebTest.Context["$LoadTestUserContext"];
...
// setting the value in the user context (i.e. in the extraction rule)
userContext["sessionId"] = "(extracted session id)";
...
// getting the value from the user context (i.e. in WebTestPlugin PreWebTest event)
e.WebTest.Context["sessionId"] = userContext["sessionId"];
You'll have to add a WebTestPlugin (that fetches the value from the user context into the web test context) to all of your web tests to make the value available across all tests.
Problem 2: Login/Logout only at start and end of load test
extract the login and logout functionality into their own separate tests (remember that the logout test also needs the WebTestPlugin that fetches the stored sessionId)
in the Load Test, the Edit Test Mix dialog lets you specify an Initialize and Terminate test: set these to the Login and Logout tests you just created
in the Load Test Scenario, set "Percentage of New Users" to 0.
Some additional explanation of the "Percentage of New Users" setting
The "Percentage of New Users" setting is poorly named and does not indicate its full behaviour.
When a "New User" starts a test iteration, it takes a new $WebTestUserId (and gets a new fresh user context, which you don't want)
When a non-"New User" starts a test iteration, it keeps the same old $WebTestUserId (and the old user context, which you do want)
So far so good. But the unexpected part is this:
Each "New User" executes the following during a load test:
Initialize > web test iteration > Terminate
A non-"New User" executes the following for the entire duration of the load test:
Initialize > iteration1 > iteration2 > ... > iterationN > Terminate
In other words, "New Users" are constantly logging in and out (which you don't want). Non-"New Users" only login and logout once in the entire load test, and continually run test iterations for the duration (which you do want).
Related
I try to call an .exe file from a webapplication.
But I want the file called by the user that is impersonalisated by windoes authentication from the website.
Process process = new Process();
try
{
process.StartInfo.UseShellExecute = false;
process.StartInfo.FileName = ConfigData.PVDToBudgetDBexePath;
process.StartInfo.CreateNoWindow = false;
process.Start();
log.Info("Process started by " + WindowsIdentity.GetCurrent().Name + " with ID: " + process.Id);
process.WaitForExit();
log.Info("After WaitForExit Process ID: " + process.Id);
}
catch (Exception ex)
{
log.Error("Error executing file with message " + ex.Message);
}
Both info log texts are logged correctly. There is no error occuring.
But the called Program does not do anything. No logging, no writing in Database.
The user has executable rights on the file.
When I call the same Code from Development Server it works fine.
I use .Net 4.5 and IIS 7
I found posts concerning this topic only for very old versions of .Net and IIS and that could not help me.
What am i doing wrong?
Or how can I find out whats going wrong?
many thanks,
EDIT:
To better make clear what I intend:
I have this (self made) exe file that imports Data from Excel Sheets into a Database. That needs some time. While doing this it logs its Progress whith log4net also into the database.
I want an UI (web application) were the user can trigger the import.
on this UI there is also an ajax progressbar that shows the progress of the import takten from the log table in the database.
I want maximum one instance of this import process to run in the same time. So I have a function that checks wheter the process is still running.
If so it does not allow to start another process. If not you can start it again.
private bool IsRunning(string name)
{
Process[] processlist = Process.GetProcesses();
if (Process.GetProcessesByName(name).Length > 0 )
{
return true;
}
else
{
return false;
}
}
I solved the problem now by starting the exe file via TimeScheduler.
path = filepath to the exe file
arguments = arguments to start the exe file with
using Microsoft.Win32.TaskScheduler;
using (TaskService taskService = new TaskService())
{
var taskDefinition = taskService.NewTask();
taskDefinition.RegistrationInfo.Author = WindowsIdentity.GetCurrent().Name;
taskDefinition.RegistrationInfo.Description = "Runs exe file";
var action = new ExecAction(path, arguments);
taskDefinition.Actions.Add(action);
taskService.RootFolder.RegisterTaskDefinition("NameOfTask", taskDefinition);
//get task:
var task = taskService.RootFolder.GetTasks().Where(a => a.Name == "NameOfTask").FirstOrDefault();
try
{
task.Run();
}
catch (Exception ex)
{
log.Error("Error starting task in TaskScheduler with message: " + ex.Message);
}
}
If you mean by development server the web server that is launched by Visual Studio, than this gives you a false test case since that server is launched by Visual Studio and uses your Windows account to run, while a standard configured IIS does not run under a "user" account but a very limited system account (luckily !!). Even if the user is logged in with a domain account in your website, the IIS process will not run under this account (that wouldn't make sense anyway). That is the reason why this code will not run in IIS and will run in your development server. Even if you get the exe to launch, it will run using the system account of IIS since you didn't supply any account, which is a limited account which will again run the exe different than you expected.
You will have to use impersonation, if you really want to go this way, but you will have to launch that process "impersonating" the user that is logged in in the website, asuuming that user account used to login even makes sense at that point. E.g. if it is a domain account, this might work, but if you use some other kind of authentication, like forms authentication, this has no meaning on OS level and thus you cannot use those credentials for impersonation in IIS.
In my experience, and I have done this a few times, impersonation in IIS is always a bad thing and is always creating issues, the same goes for launching command line process by the way.Luckily there is always a better/alternative solution when you think about it. Also the wait for a process to end in your code is not really a good practice. What if the process blocks? It will block website.
Luckily there is always a better/alternative solution when you think about it. A better/possible solution here is to use message queuing for example, where you just push a message to execute the task, and on the other end an application which processes the messages, which might use this command line tool then. That application can run under any user account you want, without you having to let IIS run under a different account. Later on you must of course come back to find the result of the operation, but that can be done using a callback in the background of your website. though this solution is a little bigger than what you are trying to do, it will have a better result on almost every field (responsiveness of your site, maintainability, scalability,..) the only thing where it is worse is the lines of code that you will need, but that is seldomly a valid factor to take into account
If you write the appplication for excel processing yourself, you can use a table in the DB as some kind of queue instead of using a message bus. Your web application then just needs to add rows with all necesarry info for the process in that table, the status and progress being one of them. Extend your processing application to monitor this table continuously and as soon as it detects a new record, it can then start to do the necessary task and update the db accordingly progress and status and end result). This avoids the messaging sub-system, will work equally good and will avoid you to have to launch a process with impersonation, which was the evil thing to start with.
You can modify the excel process to a windows service so that it runs continuously and starts with the system, but, if you don't want to, there are also tools to run any command line application as a windows service).
This technique would be much easier than the impersonation and allows your website to run in it's protected environment
How to get the user who initiated the process in IBM BPM 8.5. I want to reassign my task to the user who actually initiated the process. How it can be achieved in IBM BPM?
There are several ways to get that who initiated a Task , But who initiated a process Instance is somewhat different.
You can perform one out of the following :
Add a private variable and assign it tw.system.user_loginName at the POST of start. you can access that variable for user who initiated the process.(It will be null or undefined for the scenario if task is initiated by some REST API or UCA.)
Place a Tracking group after Start event . Add a input variable to it as username , assign it a value same as tw.system.user_loginName. So whenever Process is started entry will be inserted to DB Table.You can retrieve this value from that view in PerformanceDB.
Also there might be some table ,logging the process Instances details , where you can find the user_id directly.
I suggest you to look in getStarter() method of ProcessInstanceData API.
Official Documentation on API
This link on IBM Developerworks should help you too: Process Starter
Unfortunately there's not an Out Of The Box way to do this - nothing is recorded in the Process Instance that indicates "who" started a process. I presume this is because there are many ways to launch a process instance - from the Portal, via a Message Event, from an API call, etc.
Perhaps the best way to handle this is to add a required Input parameter to your BPD, and supply "who" started the process when you launch it. Unfortunately you can't supply any inputs from the OOTB Portal "New", but you can easilty build your own "launcher".
If you want to route the first task in process to the user that started the process the easiest approach is to simply put the start point in the lane, and on the activity select routing to "Last User In Lane". This will take care of the use case for you without requiring that you do the book keeping to track the user.
Its been a while since I've implemented this, so I can't remember if it will work elegantly if you have system steps before the first task, but this can easily be handled by moving the system steps into the human service to be executed as part of that call, rather than as a separate step in the BPD.
Define variable as string type and using script task to define the login user that use this task and assign it to your defined variable to keep to you in all of the process as initiator of the task.
You can use this line of code to achieve the same:
tw.system.user_loginName
I have a scheduled task set up to run Scan.aspx every 3 minutes in IE7. Scan.aspx reads data from 10 files in sequence. These files are constantly being updated. The values from the file are inserted into a database.
Sporadically, the value being read is truncated or distorted. For example, if the value in the file was "Hello World", random entries such as "Hello W", "Hel", etc. will be in the database. The timestamps on these entries appear completely random. Sometimes at 1:00 am, sometimes at 3:30 am. And some nights, this doesn't occur at all.
I'm unable to reproduce this issue when I debug the code. So I know under "normal" circumstances, the code executes correctly.
UPDATE:
Here is the aspx codebehind (in Page_Load) to read a text file (this is called for each of the 10 text files):
Dim filename As String = location
If File.Exists(filename) Then
Using MyParser As New FileIO.TextFieldParser(filename)
MyParser.TextFieldType = FileIO.FieldType.Delimited
MyParser.SetDelimiters("~")
Dim currentrow As String()
Dim valueA, valueB As String
While Not MyParser.EndOfData
Try
currentrow = MyParser.ReadFields()
valueA= currentrow(0).ToUpper
valueB = currentrow(1).ToUpper
//insert values as record into DB if does not exist already
Catch ex As Exception
End Try
End While
End Using
End If
Any ideas why this might cause issues when running multiple times throughout the day (via scheduled task)?
First implement a Logger such as Log4Net in your ASP.NET solution and Log method entry and exit points in your Scan.aspx as well as your method for updating the DB. There is a chance this may provide some hint of what is going on. You should also check the System Event Log to see if any other event is associated with your failed DB entries.
ASP.NET is not the best thing for this scenario especially when paired with a Windows scheduled task; this is not a robust design. A more robust system would run on a timer inside a Windows-Service-Application. Your code for reading the files and updating to the DB could be ported across. If you have access to the server and can install a Windows Service, make sure you also add Logging to the Windows Service too!
Make sure you read the How to Debug below
Windows Service Applications intro on MSDN: has further links to:
How to: Create Windows Services
How to: Install and Uninstall Services
How to: Start Services
How to: Debug Windows Service Applications]
Walkthrough: Creating a Windows Service
Application in the Component Designer
How to: Add Installers to Your Service Application
Regarding your follow up comment about the apparent random entries that sometimes occur at 1am and 3.30am: you should:
Investigate the IIS Log for the site when these occur and find out what hit(visited) the page at that time.
Check if there is an indexing service on the server which is visiting your aspx page.
Check if Anti-Virus software is installed and ascertain if this is visiting your aspx page or impacting the Asp.Net cache; this can cause compilation issues such as file-locks on the aspnet page in the aspnet cache; (a scenario for aspnet websites as opposed to aspnet web applications) which could give weird behavior.
Find out if the truncated entries coincide with the time that the files are updated: cross reference your db entries timestamp or logger timestamp with the time the files are updated.
Update your logger to log the entire contents of the file being read to verify you've not got a 'junk-in > junk-out' scenario. Be careful with diskspace on the server by running this for one night.
Find out when the App-Pool that your web app runs under is recycled and cross reference this with the time of your truncated entries; you can do this with web.config only via ASP.NET Health Monitoring.
Your code is written with a 'try catch' that will bury errors. If you are not going to do something useful with your caught error then do not catch it. Handle your edge cases in code, not a try catch.
See this try-catch question on this site.
Does webdriver maintain unique browser sessions by default when used with multiple threads i.e multiple tests in parallel? If not then how do I make it maintain unique sessions?
By
Using TestNG , we can open multiple browser session (firefox) and run tests.
I am closer to "no" - If I run my tests in Selenium Grid and some browser window "dies" (hangs up unexpectedly, because I am bad programmer), restarting the tests cause my webapp to tell me "another user with same user name is already logged in"
But in normal Selenium Webdriver, calling driver = new FirefoxDriver(); caused new session. Always.
You can use Grid configuration; It is exactly what you need.See here an example of parallel test run.
Grid can support multiple sessions. You can configure this when you register your node to the hub, using parameters : -maxSession x -browser browserName=firefox,maxInstances=x, where x represents desired number of sessions.
The Goal
I would like only a certain group of users (who are in an Active Directory group composed of users from two domains) to be able to execute a web script, in http://www.site.org/protected, after being challenged for authentication.
The Setup
Windows 2008, IIS 7. User Account Access has been disabled, as it is a pain and sometimes causes perfectly reasonable things to fail. The server is part of a domain I will called LITTLEDOMAIN. We have a trust with BIGDOMAIN.
I have a group, called "LITTLEDOMAIN\can-use-this." In that group are the members LITTLEDOMAIN\me and BIGDOMAIN\me. I did the bit in Active Directory where the server now allows that group to authenticate against another domain (BIGDOMAIN).
The application pool for www.site.org runs as "NetworkService."
The dirctory has the user SYSTEM, the user NETWORK SERVICE, the group Enterprise Admins, and the group LITTLEDOMAIN\can-use-this with at least Read and Execute permissions.
In IIS 7, I have disabled all forms of authentication for that directory but Windows Authentication. As to Authorization Rules, All Users are Allowed.
The Error
When I use, say, FireFox, visit the URL http://www.site.org/protected, and am presented with a challenge, I can enter the username LITTLEDOMAIN\me and my password, then see the minimal HTML generated by my very simple Python script, which is basically a "Hello, World" with a timestamp thrown in so I can make sure caching of the page does not occur. If I use BIGDOMAIN\me, I receive an HTTP 500 error.
Diagnostics Performed
The passwords for LITTLEDOMAIN\me and BIGDOMAIN\me are the same; this has been checked.
I look in the HTTP logs and see the "500 0 267" for "sc-status sc-substatus sc-win32-status" in the HTTP logs. A "net helpmsg 267" from the command line gives me "The directory name is invalid."
I added Failed Request Tracing Rules and see the same unhelpful message in the XML: "The directory name is invalid. (0x8007010b)"
I have turned on file object auditing in the policy for that server, then set the auditing for the directory and the files within it to have all failures for "Everyone," but nothing shows up in the Security section of Event Viewer. I was able to cause other failures, so I know that failure auditing is working. This suggests that the system is not even getting to the point where the file is being accessed.
I gave, temporarily, the group LITTLEDOMAIN\can-use-this full control over the C:\TEMP directory, on the off chance this was in use. I recycled the application pool. The same error occurs. I tried this in C:\Windows\Temp as well, to no avail.
The Question
How can I find out "well, WHICH directory name is invalid?" It's pretty obvious that something, somewhere along the line, wants permissions for BIGDOMAIN, but I cannot figure out where.
The missing component, in addition to an audit policy and Failed Request Tracing, is Process Monitor. Not Process Explorer, but Process Monitor.
Run Process Monitor for three or so seconds, just long enough to get your request in, and have it fail. Use Failed Request Tracing to get the process ID that failed. Use Process Monitor's filter to show only events where the process ID appears -- you can then see where it fails.
Then set auditing on that directory to see what account is being used.
It appears that IIS 7 is returning to the root of the webserver when looking at a protected subdirectory. Odd.