Fix XmlReader Test on ARM

[MichelZ]

DataStreamTest.ExecuteXmlReaderTest fails on ARM (and only on ARM, but both, NET462 and NET8) with:

[xUnit.net 00:02:15.08]     Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer_TCP [FAIL]
  [xUnit.net 00:02:15.08]       FAILED: Actual XML results differed from expected value.
  Expected: <orders orderid="10248" /> (System.String)
  Actual: <orders orderid="10250" /> (System.String)
  [xUnit.net 00:02:15.08]       Expected: True
  [xUnit.net 00:02:15.08]       Actual:   False
  [xUnit.net 00:02:15.08]       Stack Trace:
  [xUnit.net 00:02:15.08]         /_/src/Microsoft.Data.SqlClient/tests/ManualTests/DataCommon/DataTestUtility.cs(724,0): at Microsoft.Data.SqlClient.ManualTesting.Tests.DataTestUtility.AssertEqualsWithDescription(Object expectedValue, Object actualValue, String failMessage)
  [xUnit.net 00:02:15.08]         /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs(796,0): at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.ExecuteXmlReaderTest(String connectionString)
  [xUnit.net 00:02:15.08]         /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs(168,0): at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer(String connectionString, Boolean usingNamePipes)
  [xUnit.net 00:02:15.08]         /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs(38,0): at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer_TCP()
    Failed Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer_TCP [2 s]
##[error]EXEC(0,0): Error Message: 
EXEC : error Message:  [D:\a\_work\1\s\build.proj]
     FAILED: Actual XML results differed from expected value.
  Expected: <orders orderid="10248" /> (System.String)
  Actual: <orders orderid="10250" /> (System.String)
  Expected: True
  Actual:   False
    Stack Trace:
       at Microsoft.Data.SqlClient.ManualTesting.Tests.DataTestUtility.AssertEqualsWithDescription(Object expectedValue, Object actualValue, String failMessage) in /_/src/Microsoft.Data.SqlClient/tests/ManualTests/DataCommon/DataTestUtility.cs:line 724
     at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.ExecuteXmlReaderTest(String connectionString) in /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs:line 796
     at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer(String connectionString, Boolean usingNamePipes) in /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs:line 168
     at Microsoft.Data.SqlClient.ManualTesting.Tests.DataStreamTest.RunAllTestsForSingleServer_TCP() in /_/src/Microsoft.Data.SqlClient/tests/ManualTests/SQL/DataStreamTest/DataStreamTest.cs:line 38

The test is currently disabled with an ActiveIssue "5540", would be great to get the description for this from Azure DevOps

[David-Engel]

The issue just has the error that the test was throwing at the time:

RunAllTestsForSingleServer_TCP variation fails with this exception:
Microsoft.Data.SqlClient.SqlException : A connection was successfully established with the server, but then an error occurred during the pre-login handshake. (provider: TCP Provider, error: 0 - The specified network name is no longer available.)
2018-11-19T22:31:46.1819832Z ---- System.ComponentModel.Win32Exception : The specified network name is no longer available

In 2018, (MDS ~1.x) there was likely no time to investigate further. But given the new error, it looks like things have progressed since then.

[MichelZ]

@David-Engel I don't think so, have a look at these test results:
https://dev.azure.com/sqlclientdrivers/public/_build/results?buildId=102372&view=logs&j=c4d123bd-af38-58be-f467-a90b89592a05&t=3d22ce1c-8a2a-5bf5-a576-9c31e7813191

The tests only fail on ARM, they fail twice per Job, and only TestSet 2 fails.
So this test fails 6 times on ARM with the exact same message, but succeeds on all others

That's very suspicious to me

[MichelZ]

Ah, apologies, I missunderstood you. You've posted the description of the original Issue :)

[edwardneal]

The relevant sections of the test are here. I think this is actually a problem with the test: it expects to receive results in a specific sequence, but never specifies an ORDER BY in the query; SQL Server will thus just use the order specified in the clustered index/PK. Does the test fail if you include the ORDER BY statement?

There's a similar point to be made for the query on line 682.

[MichelZ]

The relevant sections of the test are here. I think this is actually a problem with the test: it expects to receive results in a specific sequence, but never specifies an ORDER BY in the query; SQL Server will thus just use the order specified in the clustered index/PK. Does the test fail if you include the ORDER BY statement?

There's a similar point to be made for the query on line 682.

Yes, of course! Thank you