Anith Gopal

  There’s been a lot of posts on using the Roslyn CTP APIs for syntax trees, semantic binding, and IDE smart editor features.

[I originally posted this as an editorial for the MSDN Flash ] Happy New Year! As some of you may have spotted , I’m a certified cricket tragic. In fact, I’m writing this from the top deck of the M A Noble stand, waiting for play to start on day 2 of the Sydney test.

Hola a todos, me encontre este excelente recurso, aqui el detalle: Microsoft Codename “Cloud Numerics” Model and analyze data at scale Do you need to model or analyze data that is so large it simply doesn’t fit in your desktop memory?

Live shows are always a tricky thing. On the one hand, they are great as they allow you to interact with the show, ask questions, give your opinions, etc. On the other hand, they are trickier from a technical perspective and many things can go wrong, as was the case last Wednesday when this episode was originally scheduled to air.

SQL Server 2008 中包含一个新功能，旨在帮助解决特别棘手的连接问题。这个新功能是 Connectivity Ring Buffer ，它可以捕捉每一个由服务器发起的连接关闭记录 (server-initiated connection closure) ，包括每一个 session 或登录失败事件。为了进行有效的故障排除， Ring Buffer 会尝试提供客户端的故障和服务器的关闭动作之间的关系信息。只要服务器在线 , 最高 1K 的 Ring Buffer 就会被保存， 1000 条记录后， Buffer 开始循环覆盖，即从最老的记录开始覆盖。 Connectivity Ring Buffer 的记录是能够使用 DMV 查询的：   SELECT CAST ( record AS XML ) FROM sys . dm_os_ring_buffers WHERE ring_buffer_type = ‘RING_BUFFER_CONNECTIVITY’   上述指令会选择所有记录为 XML 类型 ; 在 Management Studio 中 , 你可以单击记录 , 从而获得更具可读性的版本。如果你想使用 SQL 查询 XML 记录从而找到相应的问题 , 你可以使用 SQL server 的 XML 支持 , 将之变为一个临时的表 , 从而查询记录。   一个基本的Buffer entry：Killed SPID 一个导致服务器发起的连接关闭的简单方法是打开两个 SQL 服务器的连接，找到一个连接的 SPID ，然后从另一个连接中将该 SPID 杀死。 C:>osql -E 1> SELECT @@ spid 2> go —— 51 (1 row affected) C:>osql -E 1> kill 51 2> go 1> 如果你做了上述工作，然后查询 Ring Buffer ，你会得到和如下类似的结果： < Record id = ” 2 ” type = ” RING_BUFFER_CONNECTIVITY ” time = ” 110448275 ” > < ConnectivityTraceRecord > < RecordType > ConnectionClose </ RecordType > < RecordSource > Tds </ RecordSource > < Spid > 55 </ Spid > < SniConnectionId > B7882F3C-3BA9-45A7-8D23-3C5C05F9BDF9 </ SniConnectionId > < SniProvider > 4 </ SniProvider > < RemoteHost > & lt ; local machine &gt ; </ RemoteHost > < RemotePort > 0 </ RemotePort > < LocalHost /> < LocalPort > 0 </ LocalPort > < RecordTime > 5/6/2008 22:47:35.880 </ RecordTime > < TdsBuffersInformation > < TdsInputBufferError > 0 </ TdsInputBufferError > < TdsOutputBufferError > 0 </ TdsOutputBufferError > < TdsInputBufferBytes > 60 </ TdsInputBufferBytes > </ TdsBuffersInformation > < TdsDisconnectFlags > < PhysicalConnectionIsKilled > 0 </ PhysicalConnectionIsKilled > < DisconnectDueToReadError > 0 </ DisconnectDueToReadError > < NetworkErrorFoundInInputStream > 0 </ NetworkErrorFoundInInputStream > < ErrorFoundBeforeLogin > 0 </ ErrorFoundBeforeLogin > < SessionIsKilled > 1 </ SessionIsKilled > < NormalDisconnect > 0 </ NormalDisconnect > < NormalLogout > 0 </ NormalLogout > </ TdsDisconnectFlags > </ ConnectivityTraceRecord > < Stack > < frame id = ” 0 ” > 0X01CA0B00 </ frame > < frame id = ” 1 ” > 0X01CA0DB1 </ frame > < frame id = ” 2 ” > 0X01DF6162 </ frame > < frame id = ” 3 ” > 0X02E53C98 </ frame > < frame id = ” 4 ” > 0X02E54845 </ frame > < frame id = ” 5 ” > 0X02E57BE9 </ frame > < frame id = ” 6 ” > 0X02E38F57 </ frame > < frame id = ” 7 ” > 0X02E3B2C0 </ frame > < frame id = ” 8 ” > 0X02E3C832 </ frame > < frame id = ” 9 ” > 0X02E3D55E </ frame > < frame id = ” 10 ” > 0X781329BB </ frame > < frame id = ” 11 ” > 0X78132A47 </ frame > </ Stack > </ Record > 不同的记录类型包括不同的信息。 Connectivity Ring Buffer 记录的三种记录类型分别是： ConnectionClose ， Error ，和 LoginTimers 。上面的结果是一个 ConnectionClose ，因为这不是一个登陆时超时，或者其它的登陆失败的场景： < RecordType > ConnectionClose </ RecordType > 我们可以看出， SPID   55 的连接关闭了： <![endif]> < Spid > 55 </ Spid > 我们可以看到连接是本地的（ <local machine> 表明其是一个本地的， shared memory 类型的连接）。 <![endif]> < RemoteHost > & lt ; local machine &gt ; </ RemoteHost >   当使用 TCP 协议进行连接时，可以获得更多的相关信息 – 例如，本地 IP 地址，端口，以及远程 IP 地址和端口，从而允许你唯一的确定客户机及其应用。另外， Ring Buffer 包括了一个时间戳以及与之相对应的 SPID （如果有的话），这样才能形成一个完整的对应关系。（因为随着时间的推移 SPID 会被不同的连接所重用）。 我们同样可以看到客户发的 TDS 包中有多少 bytes ，并且可以知道是否在 TDS 中有任何的错误： < TdsInputBufferError > 0 </ TdsInputBufferError > < TdsOutputBufferError > 0 </ TdsOutputBufferError > < TdsInputBufferBytes > 60 </ TdsInputBufferBytes >   最相关的，最易于分析的信息记录在 TdsDisconnectFlags 中，有一系列的值，记录了关闭连接的状态。这里，我们看到没有发现错误，但是这里记录了这也不是一个正常的断开或者一个正常的登出。从如下的 flag 中，这个 session 是被杀死的： < SessionIsKilled > 1 </ SessionIsKilled > 一个更有意思的例子：DC 连接性问题 跟踪被杀死的 SPID 看起来很 cool 。但是 Connectivity Ring Buffer 更重要的最用 是帮助我们可以在不使用 network monitor 的情况下来解决棘手的问题。以下是一个 Connectivity Ring Buffer 中 Login Time 记录的例子，如果没有代价高昂的问题重现过程并且分析网络抓获的包，这个问题很难查明： < Record id = ” 3 ” type = ” RING_BUFFER_CONNECTIVITY ” time = ” 112254962 ” > < ConnectivityTraceRecord > < RecordType > LoginTimers </ RecordType > < Spid > 0 </ Spid > < SniConnectionId > B401B045-3C82-4AAC-A459-DB0520925431 </ SniConnectionId > < SniConsumerError > 17830 </ SniConsumerError > < SniProvider > 4 </ SniProvider > < State > 102 </ State > < RemoteHost > & lt ; local machine &gt ; </ RemoteHost > < RemotePort > 0 </ RemotePort > < LocalHost /> < LocalPort > 0 </ LocalPort > < RecordTime > 5/6/2008 23:17:42.556 </ RecordTime > < TdsBuffersInformation > < TdsInputBufferError > 0 </ TdsInputBufferError > < TdsOutputBufferError > 232 </ TdsOutputBufferError > < TdsInputBufferBytes > 198 </ TdsInputBufferBytes > </ TdsBuffersInformation > < LoginTimers > < TotalLoginTimeInMilliseconds > 21837 </ TotalLoginTimeInMilliseconds > < LoginTaskEnqueuedInMilliseconds > 0 </ LoginTaskEnqueuedInMilliseconds > < NetworkWritesInMilliseconds > 2 </ NetworkWritesInMilliseconds > < NetworkReadsInMilliseconds > 77 </ NetworkReadsInMilliseconds > < SslProcessingInMilliseconds > 3 </ SslProcessingInMilliseconds > < SspiProcessingInMilliseconds > 21756 </ SspiProcessingInMilliseconds > < LoginTriggerAndResourceGovernorProcessingInMilliseconds > 0 </ LoginTriggerAndResourceGovernorProcessingInMilliseconds > </ LoginTimers > </ ConnectivityTraceRecord > < Stack > < frame id = ” 0 ” > 0X01CA0B00 </ frame > … < frame id = ” 15 ” > 0X02E3C832 </ frame > </ Stack > </ Record > 在这个情况下，在客户端，我们可以看到： [SQL Server Native Client 10.0]Shared Memory Provider: Timeout error [258]

Just like it says in the title. Whitepaper here .

Agile in Action 1 : Planning in Agile 진정한 Agile 이란 무엇일까요?  본 세션에서는 Agile 프로세스, Lean 개발과 ALM 등에 대한 진정한 의미와 그 중요성에 대해서 설명해 드릴 것입니다. 마이크로소프트의 ALM 을 바라보는 관점과 Agile 을 위해 어떠한 개발플랫폼을 지원하는 지 알아보세요

Following this recent announcement last month from the System Center Service Manager team: New SCSM–Microsoft Project Connector Released , I finally got a chance to test out this very nice solution built by Expit’s . In their own word: Expit has developed a plugin/connector that enables the exchange of information between project planning software (Project server 2010) and System Center Service Manager. The connector allows project server users to convert project tasks into tickets in service manager directly from project server or project professional clients that are connected to project server.

  http://joelcitizen.com/and-so-it-will-someday Joel Reyes

Once again we’ve captured all of the main track sessions from TechEd Australia this year. We grabbed whatever video went to the screen and whatever audio went to the speakers.

오늘 소개해 드릴 내용은 XBOX용으로 출시된 Kinect 입니다.

El día de hoy se liberó la actualización del Visual Studio 2010 and .NET Framework 4 Training Course , aquí los detalles: Visual Studio 2010 and .NET Framework 4 Training Course Units   Overview In this introductory unit, the Visual Studio 2010 and .NET Framework 4 evangelism team welcomes you to the training course and goes over the significant changes to Visual Studio 2010 and the .NET Framework 4. We’ll also cover what you’ll see in the course and how to follow along with the labs.   Managed Languages There are same great enhancements to the set of managed languages in Visual Studio 2010.