Develop Reference

Multiple NUnit tests reading in a long stream

I'm new into unit testing and I came across a problem.
My application is reading in a file with lots of data frames and the tests are supposed to validate how my program deals with them.
There are different kind of errors that can occur reading these frames, and they can happen anywhere in the file. Reading through the entire file can take 2 seconds and the amount of tests can become massive.
Which means that reading in the file just once and testing it for all errors in that one run would be the most effective way. However, that is against the common test pattern I see recommended everywhere. Common test pattern recommend to have 1 test method per test, but this would mean if I want to test 10 things I have to read through the entire file 10 times. Especially when they get more and more this would be very ineffective through. There are no shortcuts possible, all these tests require to read the whole file.
I know the types of errors that can occur, I just do not know where and if it's there before the test starts.
My current "dirty" solution is to add all errors to a list and output that list later with an assert.fail, but I've read multiple times that this is indeed a very dirty solution.
One idea I had was that one thread reads through the file and "feeds" all test methods with the next record, and waits for them until it proceeds. However, I do not believe it is possible to pull this off.
Any idea on how to make this clean and effective? Or should I just take the cheat and add them all to a list like I'm doing right now? Most internet examples are too much perfect world examples and it's hard to find solutions for things like this.
Edit: I'm aware it's possible to execute tests in Parallel, but wouldn't this just delay the inevitable bottle neck?
Edit2:
This is the design of the code.
private List<TestError> errors;
private Session session;
private MemoryStream mStr;
private void Init()
{
errors = new List<TestError>();
session = new Session();
var test = NUnit.Framework.TestContext.CurrentContext.Test;
mStr = new MemoryStream(File.ReadAllBytes(TestContext.CurrentContext.TestDirectory + "\\Files\\" + test.ClassName.Split('.')[2] + "_" + test.MethodName + ".bin"));
session.StartRawDataRecording(mStr,
true, true);
}
[Test]
public async Task Test1()
{
Init();
var recorder = session.Recorder as RawDataRecorder;
while (!recorder.IsCancelled)
{
await recorder.ManualLoopTick();
if({condition1})
errors.Add(new TestError("condition1", "condition1 occured"));
if({condition2})
errors.Add(new TestError("condition2", "condition2 occured"));
}
if(errors.Count > 0)
{
StringBuilder b = new StringBuilder();
b.AppendLine("Run Errors:");
for (int i = 0; i < errors.Count;++i)
{
var err = errors[i];
b.AppendLine(String.Format("Error: {0}, Message: {1}",err.Error,err.Message));
b.AppendLine("Stacktrace: " + err.StackTrace);
b.AppendLine();
}
Assert.Fail(b.ToString());
}

System.Windows.Automation is very slow at enumerating table rows vs. UIAutomationCore

I am trying to do automated testing of my application via UI Automation (mainly using TestStack.White to provide a friendly interface; it uses System.Windows.Automation as a back-end). I have a table with ~200 rows that I need to test the values of (actually I only want to test the first and last couple rows). I have discovered that using COM-interop UIAutomationCore by itself, I can enumerate the rows in a fraction of a second, but only when I don't use White or System.Windows.Automation. As soon as System.Windows.Automation initializes, future UI Automation actions to enumerate rows are slow:
First COM run: it took 0.04 seconds to get 102 rows!
First System.Windows.Automation run: it took 7.18 seconds to get 102 rows!
Second COM run: it took 7.87 seconds to get 102 rows!
I created a simple WinForms test application (TableTest.exe to verify that it was System.Windows.Automation and not something to do with my application:
static void Main()
{
Application.EnableVisualStyles();
Application.SetCompatibleTextRenderingDefault(false);
var form = new Form() { Text = "TableTest", WindowState = FormWindowState.Maximized };
var dgv = new DataGridView() { Name = "DGV", Dock = DockStyle.Fill, AutoSizeColumnsMode = DataGridViewAutoSizeColumnsMode.Fill };
dgv.Columns.Add("i", "i");
dgv.Columns.Add("2i", "2i");
dgv.Columns.Add("i^2", "i^2");
dgv.Columns.Add("i^i", "i^i");
for (int i = 0; i < 100; ++i)
dgv.Rows.Add(i, i * 2, i * i, Math.Pow(i, i));
form.Controls.Add(dgv);
Application.Run(form);
}
Then I created another test app to test the first one. It works as either a console app or a WinForms app. First I test with COM automation, then with System.Windows.Automation, then again with COM automation. As you can see from the output I quoted above, the first block executes very quickly, the next two blocks execute excruciatingly slowly. If I comment out the System.Windows.Automation block code then both COM blocks execute quickly.
using UIA = Interop.UIAutomationCore;
static void Main(string[] args)
{
var process = System.Diagnostics.Process.Start("TableTest.exe");
System.Threading.Thread.Sleep(500);
var uia = new UIA.CUIAutomation();
var rootCom = uia.GetRootElement();
var windowCom = rootCom.FindFirst(UIA.TreeScope.TreeScope_Children, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_NamePropertyId, "TableTest"));
var dgvCom = windowCom.FindFirst(UIA.TreeScope.TreeScope_Descendants, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_AutomationIdPropertyId, "DGV"));
var start = DateTime.Now;
var rowCount = dgvCom.FindAll(UIA.TreeScope.TreeScope_Children, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_ControlTypePropertyId, UIA.UIA_ControlTypeIds.UIA_CustomControlTypeId)).Length;
var elapsed = (DateTime.Now - start).TotalSeconds;
Console.WriteLine(String.Format("It took {0} seconds to get {1} rows!", elapsed.ToString("f2"), rowCount));
process.Kill();
process = System.Diagnostics.Process.Start("TableTest.exe");
System.Threading.Thread.Sleep(500);
var root = AutomationElement.RootElement;
var window = root.FindFirst(TreeScope.Children, new PropertyCondition(AutomationElement.NameProperty, "TableTest"));
var dgv = window.FindFirst(TreeScope.Descendants, new PropertyCondition(AutomationElement.AutomationIdProperty, "DGV"));
start = DateTime.Now;
rowCount = dgv.FindAll(TreeScope.Children, new PropertyCondition(AutomationElement.ControlTypeProperty, ControlType.Custom)).Count;
elapsed = (DateTime.Now - start).TotalSeconds;
Console.WriteLine(String.Format("It took {0} seconds to get {1} rows!", elapsed.ToString("f2"), rowCount));
process.Kill();
process = System.Diagnostics.Process.Start("TableTest.exe");
System.Threading.Thread.Sleep(500);
uia = new UIA.CUIAutomation();
rootCom = uia.GetRootElement();
windowCom = rootCom.FindFirst(UIA.TreeScope.TreeScope_Children, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_NamePropertyId, "TableTest"));
dgvCom = windowCom.FindFirst(UIA.TreeScope.TreeScope_Descendants, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_AutomationIdPropertyId, "DGV"));
start = DateTime.Now;
rowCount = dgvCom.FindAll(UIA.TreeScope.TreeScope_Children, uia.CreatePropertyCondition(UIA.UIA_PropertyIds.UIA_ControlTypePropertyId, UIA.UIA_ControlTypeIds.UIA_CustomControlTypeId)).Length;
elapsed = (DateTime.Now - start).TotalSeconds;
Console.WriteLine(String.Format("It took {0} seconds to get {1} rows!", elapsed.ToString("f2"), rowCount));
process.Kill();
}
What the heck is System.Windows.Automation doing that kills the performance of UI Automation? I've looked at the White source code and I don't see anything obvious. I can't profile System.Windows.Automation itself because I can't find any PDB for it. I'm not very familiar with UI Automation so maybe it'll be obvious to someone else. The White is: 0.13.0.0 and I'm testing on 64-bit Windows 7.

I cannot answer your question. But many people will come here from Google who search for the keywords "uiautomation slow" and the first result in Google is your question. (You wrote a bestseller)
For all those coming from Google and struggling with slow UIAutomation I post this answer.
System.Windows.Automation is EXTREMELY slow. Obtaining 30 child elements may take 1000ms on a very fast computer! I have even seen it hanging forever while getting the child elements of a Tree in a QT application.
Apart from that the implementation is not even thread safe.
System.Windows.Automation is deprecated. Do not use it!
In the MSDN you find the following note:
UI Automation was first available in Windows XP as part of the
Microsoft .NET Framework. Although an unmanaged C++ API was also
published at that time, the usefulness of client functions was limited
because of interoperability issues. For Windows 7, the API has been
rewritten in the Component Object Model (COM).
Although the library functions introduced in the earlier version of
UI Automation are still documented, they should not be used in new
applications.
The solution to slow performance is to use the new IUIAutomationElement COM interface instead of the old System.Windows.Automation C# interface. After that the code will be running lightning fast!
Apart from that the new interface offers much more patterns and Microsoft is extending it continously. In the Windows 10 SDK (UIAutomationClient.h and UIAutomationCore.h) several patterns and properties have been added which are not available in the .NET Automation framework.
The following patterns are available in the COM version of UIAutomation which do not exist in System.Windows.Automation:
IUIAutomationLegacyIAccessiblePattern
IUIAutomationObjectModelPattern
IUIAutomationAnnotationPattern
IUIAutomationTextPattern2
IUIAutomationStylesPattern
IUIAutomationSpreadsheetPattern
IUIAutomationSpreadsheetItemPattern
IUIAutomationTransformPattern2
IUIAutomationTextChildPattern
IUIAutomationDragPattern
IUIAutomationDropTargetPattern
IUIAutomationTextEditPattern
IUIAutomationCustomNavigationPattern
Additionally the following Control types have been added:
AppBar
SemanticZoom
Additionally the following Element's have been added:
IUIAutomationElement2
IUIAutomationElement3
IUIAutomationElement4

The examples you posted do not use White... FWIW, White uses recursive calls to Automation.Find requesting the children each time. This returns valid results but is slower than requesting the subtree from the appropriate parent node - note that the root node is never the 'appropriate' node from which to request the subtree (see the note on MSDN). Since your example only requests the children once, that is not the issue.

neo4j REST API poor performance

According to my benchmark of creating nodes using
GraphClient.Create()
performance leaves much to be desired.
I've got about 10 empty nodes per second on my machine (Core i3, 8 GB RAM).
Even when I use multithreading to perform create time to each Create() call speed icreases linearly (~N times when used N threads).
I've tested both stable 1.9.2 and 2.0.0-M04. The results exactly the same.
Does anybody know what's wrong?
EDIT: I tried to use neo4j REST API and I got similar results: ~ 20 empty nodes per second and multithreading also gives no benefits.
EDIT 2: At the same time Batch REST API, that allows batch creations provides much better performance: about 250 nodes per second. It looks like there is incredible big overhead in handling single request...

Poor performance caused by overhead in processing RESTful Cypher query. Mostly it is network overhead but overhead caused by need to parse query also exists.
Use Core Java API when you interested in high performance. Core Java API provides more than 10 times faster requests processing than Cypher query language.
See this articles:
Performance of Graph Query Languages
Get the full neo4j power by using the Core Java API for traversing
your Graph data base instead of Cypher Query Language

The neo4jclient itself uses the REST API, so you're already limited in performance (by bandwidth, network latency etc) when compared to a direct API call (for which you'd need Java).
What performance are you after?
What code are you running?
Some initial thoughts & tests to try:
Obviously there are things like CPU etc which will cause some throttling, some things to consider:
Is the Neo4J server on the same machine?
Have you tried your application not through Visual Studio? (i.e. no debugging)
In my test code (below), I get 10 entries in ~200ms - can you try this code in a simple console app and see what you get?
private static void Main()
{
var client = new GraphClient(new Uri("http://localhost.:7474/db/data"));
client.Connect();
for (int i = 0; i < 10; i++)
CreateEmptyNodes(10, client);
}
private static void CreateEmptyNodes(int numberToCreate, IGraphClient client)
{
var start = DateTime.Now;
for (int i = 0; i < numberToCreate; i++)
client.Create(new object());
var timeTaken = DateTime.Now - start;
Console.WriteLine("For {0} items, I took: {1}ms", numberToCreate, timeTaken.TotalMilliseconds);
}
EDIT:
This is a raw HttpClient approach to calling the 'Create', which I believe is analagous to what neo4jclient is doing under the hood:
private async static void StraightHttpClient(int iterations, int amount)
{
var client = new HttpClient {BaseAddress = new Uri("http://localhost.:7474/db/data/")};
for (int j = 0; j < iterations; j++)
{
DateTime start = DateTime.Now;
for (int i = 0; i < amount; i++)
{
var response = await client.SendAsync(new HttpRequestMessage(HttpMethod.Post, "cypher/") { Content = new StringContent("{\"query\":\"create me\"}", Encoding.UTF8, "application/json") });
if(response.StatusCode != HttpStatusCode.OK)
Console.WriteLine("Not ok");
}
TimeSpan timeTaken = DateTime.Now - start;
Console.WriteLine("took {0}ms", timeTaken.TotalMilliseconds);
}
}
Now, if you didn't care about the response, you could just call Client.SendAsync(..) without the await, and that gets you to a spiffy ~2500 per second. However obviously the big issue here is that you haven't necessarily sent any of those creates, you've basically queued them, so shut down your program straight after, and chances are you'll have either no entries, or a very small number.
So.. clearly the code can handle firing x thousand calls a second with no problems, (I've done a similar test to the above using ServiceStack and RestSharp, both take similar times to the HttpClient).
What it can't do is send those to the actual server at the same rate, so we're limited by the windows http stack and / or how fast n4j can process the request and supply a response.

determining the current link speed of WiFi in C#

I am writing a program that does one thing, it finds out the current link speed of the wifi connection and reports it to the user in real time. the problem I am having is that it does not seem to be able to find out the current link speed, only the max link speed of the device (300 Mbps). the reason I am writing this is that I have a problem where, periodically the link speed will drop drastically (down to 1-2 Mbps) and I want to be able to see when that happens. with this code it will simply give me the maximum speed that the adapter supports, not the current link speed of the connection.
private void update(object state)
{
System.Net.NetworkInformation.NetworkInterface[] nics = null;
nics = System.Net.NetworkInformation.NetworkInterface.GetAllNetworkInterfaces();
long speed = 0;
string adapter = "";
foreach (System.Net.NetworkInformation.NetworkInterface net in nics)
{
if (net.Name.Contains("Wireless") || net.Name.Contains("WiFi") || net.Name.Contains("802.11") || net.Name.Contains("Wi-Fi"))
{
speed = net.Speed;
adapter = net.Name;
break;
}
}
string temp;
if (speed == 0)
{
temp = "There is currently no Wi-Fi connection";
}
else
{
temp = "Current Wi-Fi Speed: " + (speed / 1000000) + "Mbps on " + adapter;
}
if (label1.InvokeRequired)
{
SetTextCallback d = new SetTextCallback(update);
label1.Invoke(d, new object[] { temp });
}
else
{
label1.Text = temp;
}
}
this run by calling
System.Threading.Timer ticker = new System.Threading.Timer(update, label1, 0, 1000);
in the main method.

Considering that it literally took me the whole entire day to find what the solution to this was, I figured I'd at least show StackOverflow for future reference what I came across and what did and did not work for this question.
tl;dr: Scroll to the The Code section
What I found
Good ol' control panel
If you are looking for the really easy way to do this you can simply go and open Contol Panel. Depending on what version of Windows you are on (in my case I'm on Windows 8), the path to the page is Control Panel >> Network and Internet >> Network and Sharing Center and then you can click on the link next to "Connections: " which will give you a window that looks like what is below.
The current link speed is highlighted in red which in my case is 36.0 Mbps. Though, of course, this might not satisfy your original question if you were intending to integrate some code with the actual value.
WMI
With a mix of Googling and whatnot, I thought I might have found something in Windows Management Instrumentation.
Long story short, AFAIK, WMI does not have what we're looking for.
WMI, in short, is a giant object database (that
can also be queried through SQL) that allows you to query information about a
Windows machine such as process, disks, etc. In WMI, everything is
represented by a class with a series of instances each with a set of
properties.
Anyhow, WMI Explorer allows you to view all of this on your machine.
I (supposedly) found two classes on MSDN that might have the info on link speed but from WMI Explorer, there was nothing useful.
The first class, MSFT_NetAdapter, did not even show up in WMI Explorer on my machine.
The second class, Win32_NetworkAdapter, showed up in WMI Explorer, but the Speed property was still incorrect. The same network adapter was showing a value of 168000000 or 168 Mbps which is not right. Though I find this strange because there was already a MaxSpeed but it was blank.
Scratch WMI off the list.
Win32 P/Invoke
Yes, of course, the solution to everything is always calling unmanaged Win32 APIs using P/Invoke magic.
This is the route used to solve the problem.
Luckily, the IP_ADAPTER_ADDRESSES structure solves the problem. If you look at the MSDN page, it's a fairly large structure but what is important here is TransmitLinkSpeed which actually works.
Calling the GetAdaptersAddresses() function will return the actual structure.
Now, the actual C# P/Invoke code. Luckily, pinvoke.net already had interop for this function which I've added. This is all that was necessary.
The Code
Finally, here is your code patched up with the new P/Invoke black magic. I've made it work as a console application for demo purposes:
Using Statements:
using System;
using System.Threading;
Code:
class Program
{
private static void Main(string[] args)
{
Timer ticker = new Timer(Update, null, 0, 1000);
// Keep the main thread from dying
while (true)
{
Thread.Sleep(1000);
}
}
private static void Update(object state)
{
ulong speed = 0;
string adapter = "";
string[] nameSearches = { "Wireless", "WiFi", "802.11", "Wi-Fi" };
// The enum value of `AF_INET` will select only IPv4 adapters.
// You can change this to `AF_INET6` for IPv6 likewise
// And `AF_UNSPEC` for either one
foreach (IPIntertop.IP_ADAPTER_ADDRESSES net in IPIntertop.GetIPAdapters(IPIntertop.FAMILY.AF_INET))
{
bool containsName = false;
foreach (string name in nameSearches)
{
if (net.FriendlyName.Contains(name))
{
containsName = true;
}
}
if (!containsName) continue;
speed = net.TrasmitLinkSpeed;
adapter = net.FriendlyName;
break;
}
string temp;
if (speed == 0)
{
temp = "There is currently no Wi-Fi connection";
}
else
{
temp = string.Format("Current Wi-Fi Speed: {0} Mbps on {1}", (speed / 1000000.0), adapter);
}
Console.WriteLine(temp);
}
}
You are then going to be looking for the actual IPIntertop class that I updated. Since it's pretty big you can find it updated at pinvoke.net or on this PasteBin in case something goes down.
Bottom Line
Windows has a lot of APIs which are somewhat broken (WMI), can have a few "leaky abstractions" (.Net), or can be a pain to work with (Win32).
Sigh, that is a lot and I hope it helps.

I come accross the same issue, and need to get windows wifi link speed which is current negotiated.
and thanks to #Jaxrtech's WMI approach, that really works.
the correct class is CIM_NetworkAdapter(i'm using windows7), and query the speed column to get the current speed.
while wifi current negotiated speed is changing, this speed is changing too. i tested it, this matched ok.
select Description , DeviceID, Speed from CIM_NetworkAdapter
get:
D-Link DWA-140 RangeBooster N USB Adapter 17 285000000

Since no-one here mentioned it yet: why not use https://learn.microsoft.com/en-us/dotnet/api/system.net.networkinformation.networkinterface.speed?view=net-5.0#System_Net_NetworkInformation_NetworkInterface_Speed
This is, if the table on this site is correct included since .NET Framework 2.0 and seem to included in all other versions of .net including .net core.

Trying to run multiple HTTP requests in parallel, but being limited by Windows (registry)

I'm developing an application (winforms C# .NET 4.0) where I access a lookup functionality from a 3rd party through a simple HTTP request. I call an url with a parameter, and in return I get a small string with the result of the lookup. Simple enough.
The challenge is however, that I have to do lots of these lookups (a couple of thousands), and I would like to limit the time needed. Therefore I would like to run requests in parallel (say 10-20). I use a ThreadPool to do this, and the short version of my code looks like this:
public void startAsyncLookup(Action<LookupResult> returnLookupResult)
{
this.returnLookupResult = returnLookupResult;
foreach (string number in numbersToLookup)
{
ThreadPool.QueueUserWorkItem(lookupNumber, number);
}
}
public void lookupNumber(Object threadContext)
{
string numberToLookup = (string)threadContext;
string url = #"http://some.url.com/?number=" + numberToLookup;
WebClient webClient = new WebClient();
Stream responseData = webClient.OpenRead(url);
LookupResult lookupResult = parseLookupResult(responseData);
returnLookupResult(lookupResult);
}
I fill up numbersToLookup (a List<String>) from another place, call startAsyncLookup and provide it with a call-back function returnLookupResult to return each result. This works, but I found that I'm not getting the throughput I want.
Initially I thought it might be the 3rd party having a poor system on their end, but I excluded this by trying to run the same code from two different machines at the same time. Each of the two took as long as one did alone, so I could rule out that one.
A colleague then tipped me that this might be a limitation in Windows. I googled a bit, and found amongst others this post saying that by default Windows limits the number of simultaneous request to the same web server to 4 for HTTP 1.0 and to 2 for HTTP 1.1 (for HTTP 1.1 this is actually according to the specification (RFC2068)).
The same post referred to above also provided a way to increase these limits. By adding two registry values to [HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Internet Settings] (MaxConnectionsPerServer and MaxConnectionsPer1_0Server), I could control this myself.
So, I tried this (sat both to 20), restarted my computer, and tried to run my program again. Sadly though, it didn't seem to help any. I also kept an eye on the Resource Monitor while running my batch lookup, and I noticed that my application (the one with the title blacked out) still only was using two TCP connections.
So, the question is, why isn't this working? Is the post I linked to using the wrong registry values? Is this perhaps not possible to "hack" in Windows any longer (I'm on Windows 7)?
And just in case anyone should wonder, I have also tried with different settings for MaxThreads on ThreadPool (everything from 10 to 100), and this didn't seem to affect my throughput at all, so the problem shouldn't be there either.

It is matter of ServicePoint. Which provides connection management for HTTP connections.
The default maximum number of concurrent connections allowed by a ServicePoint object is 2.
So if you need to increase it you can use ServicePointManager.DefaultConnectionLimit property. Just check the link in MSDN there you can see a sample. And set the value you need.

For quicker reference for someone. To increase the connection limit per host you can do this in your Main() or anytime before you begin making the HTTP requests.
System.Net.ServicePointManager.DefaultConnectionLimit = 1000; //or some other number > 4

Fire and forget this method from your main method. Icognito user is correct, only 2 threads are allowed to play at the same time.
private static void openServicePoint()
{
ServicePointManager.UseNagleAlgorithm = true;
ServicePointManager.Expect100Continue = true;
ServicePointManager.CheckCertificateRevocationList = true;
ServicePointManager.DefaultConnectionLimit = 10000;
Uri MS = new Uri("http://My awesome web site");
ServicePoint servicePoint = ServicePointManager.FindServicePoint(MS);
}

For Internet Explorer 8:
Run Registry Editor and navigate to following key HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Internet Explorer\MAIN\FeatureControl\FEATURE_MAXCONNECTION SPERSERVER
and
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Internet Explorer\MAIN\FeatureControl\FEATURE_MAXCONNECTION SPER1_0SERVER
If FEATURE_MAXCONNECTIONSPERSERVER and FEATURE_MAXCONNECTIONSPER1_0SERVER are missing then create them. Now create DWORD Value called iexplore.exe for both sub keys (listed above) and set their value to 10 or whatever number desired.