How limitation can be source of goodness - NoSQL, REST and more

[Level C3]

It just dawned on me. I came to a startling realisation that limitation/restriction/constraint  - which are words with negative connotation - can/will generate creativity and lead to goodness. This is more or less is saying "less is more". But looking at it from another angle.

Story of twitter

I do not know about you but I think Twitter is one of the biggest inventions of recent decades, somewhere along the lines of Gutenberg's printing or Priestly's Soda. Regardless of what most think about the revolution of social networking, I believe Twitter is not of the same breed - it is not a compressed facebook.

Twitter is centred around a stupidly simple idea: you have 140 characters to express yourself. No more. Yes, it also has re-tweet, follow, favourite, etc but these are features, if you remove them twitter will be more or less twitter although its usefulness will be limited. But if you remove the 140 character limitation suddenly it is not twitter anymore. Twitlonger is not twitter. With all due respect to @daltonc app.net with a limitation much different from 140 characters will not make a new twitter.

But why? Because by limiting you to use only 140 characters, you are forced to express yourself more succinctly. Makes you think really hard about what you want to say and remove all that matters less. Limitation leads you towards the ethos of twitter. You cannot write a line in the terms and condition of twitter  asking users to write only intelligent tweets but twitter has achieved this by enforcing its 140 characters limitation.

Why NoSQL matters

They say behind every successful man, there is a powerful woman. And I would say behind every inflexible and crippling architecture in the enterprise there is a big legacy database. Database becomes legacy soon after the first release, since other layers change but database cannot keep up. What I mean by database is an RDBMS database (SQL Server, Oracle, you name it.

We have cut down our processes, we do agile, we do lean, we have got continuous integration, we do unit testing and TDD, we do BDD and continuous deployment, we do ... we have built a process to minimise the risk and impact of change. Yet it is so difficult to change and the hardest to change is database.

The crux of this issue goes to the fact that business logic creeps into database. I recently witnessed how a complex calculation had to be done in an inaccurate rounding since it had to match the database calculation logic - yes there was calculation in database. 

We all believe that we should not put the business logic in database. But why do we keep doing it? Because we can. The problem is best practices and disciplines are difficult to enforce when project is late, we have a critical bug to fix or we need to make a quick change to the system and easiest solution is to change the stored procedure.

As far as SQL Server is concerned, we can have intra-model logic (calculated fields), domain-wide logic (stored procedure and user-defined functions) and cross-boundary logic (service broker) - and you can even deploy compiled code (SQL CLR). And since we can, we will.

And here is why I think NoSQL is useful - apart from all the hype around it. You simply cannot put business logic in it, so you don't. And this will lead into better design. So I like NoSQL not because of what I can do but because of what I cannot do. I would agree with Stonebraker's article and believe that NoSQL technologies could be low-tech compared to hi-tech RDBMS, but I cannot abuse them in the same way. NoSQL fully focuses on storage and retrieval and not replicating all those features that should not be implemented in a database (XML Manipulation, Message Bus, logic etc)

A table in SQL can translate to 5-6 Redis data structures so that you could effectively query and access data. So there is more work to be done but I like it since I would really think about what I need to store and what I need to query back (remember twitter?).

Lessons from REST

Regardless of all the bloated hype and endless controversies in interpreting REST, it works. It just simply works. So we have a set of constraints and if you follow them it will lead to goodness. Example? HTTP.

REST is an absolute example of how constraints can lead to a better design.

Limitation in creative arts

I am big fan of minimalism and Philip Glass is one of my favourite contemporary composers. In minimalism, a compact musical idea is repeated to create rhythm, melody and ultimately harmony - and this is usually created using layers of repetitive chords. For me who love minimal music, Satyagraha opera is the pinnacle of minimal musical expression through a limiting set of musical material. 

Apart from musical material, limitation in number of instruments is also an important aspect. String quartet is one of the most expressive forms of music - and I just love it be it Beethoven or Shostakovich. Rock counterpart for the string quartet is probably the rock trio (singer as an instrument?) where some of the best music ever produced (from Jimi Hendrix and Cream to Rage Against The Machine and Nirvana). 

In social and political terms, we have experienced an explosion of modern and beautiful art in the eastern bloc during the oppression of communist governments. Composers and film directors had to find their own language to express their art. Since they could no longer look outside for inspiration, they turned inside and a new era of creativity and great art flourished: Andrei Tarkovsky, Istvan Szabo, Andrzej Wajda, Shepitko and many more. Shostakovich arguably produced his best works during the fierce Stalinist oppression. It is interesting that when the restrictions are lifted, artist is no longer able to produce the same quality of works. Wajda's Iron Man seemed like just a shabby copy of Marble Man. And Tarkovsky's two last films made outside Russia did not feel like the previous ones. 

The same oppression created the new wave of Iran Cinema with the likes of Makhmalbaf, Mehrjui, Kiarostami and others.

Now should we create an oppressive government so that we get a great artistic output?! No, but perhaps we can have a 60's style drug revolution which does the same :)

Media type: how much can you cram into a single token?

[Level C4]

Introduction

This post discusses the problems associated with the use of a single token as media type (usually as the main value of the Content-Type header in HTTP response or Accept header in request) to describe all attributes of the content.

Motivation and background

This has been bugging me for a while. But recently I engaged in a discussion on twitter with Glenn Block @gblock and the rest of the REST enthusiast community on the options in versioning RESTful services. There are generally 2 camps: those advocating using Content Negotiation for versioning (putting version number in Content-Type header) and those preferring to stick to classic resource based versioning (including version number in the URL). Regardless of which one is better, MediaType lacks the richness required to express a media type and adding version information to a media type is not possible considering current status of the media type.

One of the main problems associated with the use of media type is its current implementation in various systems is key based, i.e. it involves matching all or none of the media type. As we will see this causes considerable problems in effective consumption of media types.

Media Type

Media type has been described in various RFCs (main one being RFC 2046) while historically these have been limited what is known as MIME types. RFC 4288 defines the procedure for registering the media types describing a formal process which needs to be followed to publicly register.

Registering a media type for a public API is all well and good but as described by this book, use of private APIs far exceeds use of public ones and registering all media types exposed within private APIs is impractical and unwarranted.

Also with popularity of REST-based APIs, there are going to be more and more service endpoints exposed. If all such services are to define new media types, we would have an explosion of media types rendering current implementation of content negotiation 

Media type is a case of an extreme semantic mix-up. A single token has been used to express many different facets of a media type. In fact the semantic space with all its axes will contain many useful points yet industry currently uses a very sparse set of points defined as media type values. Rest of this space is unusable - as such a very inefficient solution.

We will now have a look at facets/axes.

1- Human-illegibility

This is the lowest and least specific level of semantic definition of a media type. It is very simple: content of a media type can be read by a human (for example text/plain, application/xml or application/json) or the data is meant for the machine comprehension or rendering (for example image/png or video/mpeg)

Having this information separate to the actual media type can help tools such as Fiddler to decide whether they can display text of the content whose media type is unknown to the tool. Media types initially used "text" to denote such information (e.g. text/xml or text/javascript) but these have been replaced with 

2- Formatting

This is the most common and important axis of a media type information which informs the tools/clients which parser/interpreter/renderer to use for consuming such content. text/plain, application/xml, application/json, image/png or video/mpeg are all examples of such use of the media type. 

There are several known vendor-specific media types in this space such as application/vnd.ms-excel.

3- Schema

This is a further specialisation of the formatting. Common examples include application/rss+xml or application/hal+json. Basically these mean that in terms of formatting, they are the same as their parent (application/xml or application/json) yet they follow a superset schema. Use of + sign - as far as I know - is not canonical and is merely a convention followed by the industry to add schema to the established formats. Comprehension of this convention would be crucial to correct interpretation of the media type without the need for having a dictionary of all possible values, however, I believe most tools we have at the moment lack such features.

4- Domain/Vendor specific

This is where we see most of the expansion in the media type space. Basically you could output your own media type via your private API. Since you will be the main consumer of the API, integration could be easy but it is very common for private APIs to go public - especially if they are successful. An example of such media types can be found here.

5- Versioning

Versioning is the highest aspect of a media type which is normally added to Domain-specific media types. This is a popular solution to the Web API versioning problem.

For example, you could have application/mydomain.customer.1.1 as opposed to application/mydomain.customer or application/mydomain.customer.1.0

So where is the problem?

Basically information gets lost.

First problem is that clients might be interested in a lower order of these aspects of media type while in order to consume the resource, they are forced to comprehend higher order and extract the axes they are interested in. For example, a tool such as fiddler could be only interested in only whether it could display the information for the end user as plain text. A client capable of consuming XML and deserialising to objects is only interested at knowing whether it is XML while it might be represented with a media type which is essentially XML but has a different value. On the other hand, if a server uses HAL to send domain objects/view models to the client, either it has to use the standard application/hal+json or use the domain level name of the media type (with or without a version).

Another problem is that the content negotiation process will become more complex. In the lack of a standard in defining multi-axial media types, most systems implement a dictionary based rule on content negotiation as such maintaining list of possible content types becomes a burdensome task.

A solution

Basically I believe we can solve this by keeping the common media types but use media type extensions in the Content-Type header (or in the Accept header). For example:

Content-Type: application/xml; human-illegible=true; domain-name=customer; domain-version=1.1

This will ensure that existing clients and servers will not break while new clients and servers can use new extensions for content negotiation and more loosely coupled resource consumption. I will try to expand upon this idea in another post.

Conclusion

Cramming as much as information into a single token and then try parsing that one token is not a good idea especially when it comes to media type which is the communication bridge between loosely coupled world of HTTP clients and servers.

Media type token value covers 5 different aspects of the resource and separating the concerns of breaking these aspects into their own tokens can result in more robust and decoupled systems.

How PayPal is helping Iranian government's internet censorship

[Level N]

I have difficulty believing what I read in Hacker News around the same time my account was blocked and closed by PayPal trying to pay 3$ for internet proxy to combat filtering which has rendered internet pretty much useless in Iran. Did he read my email? I do not know but it left me frustrated and hopeless in one of the most difficult times in my life.

As some of you might know, I have been going through rough times for the last 18 months. My mother in law   passed away 3 weeks ago after a long battle during which my wife spent mostly with her - which is a consolation.

So when I visited her and family in Iran (that's where I am from) around 2 months ago, I realised that internet censorship is so bad that it has become really unusable. Sites such as twitter which I am addicted to are obviously blocked as they played an important role in Iran's suppressed Green Revolution, arguably first Twitter Revolution in world . In fact even this blog that you are reading and anything hosted on Google's blogspot is filtered - Iran had highest number of bloggers in the Middle East and a number of them are in prison. But even Gmail gets its share and gets blocked from time to time. According to some reports, 40% of Iranians (30 million) use internet which is second in Middle East after Israel, and as far as I know, most of them use either internet proxies, anti-filters, anonymisers and VPNs to bypass the censorship. So even if we say only half use anti-censorship tools, we are talking about 15 million people. I had actually setup a dedicated PC in UK for my wife to use as remote desktop connection (RDP) since sometimes these proxies are found by the government and blocked. But slow speed of internet (intentionally kept low) makes it almost useless since each refresh of the screen takes a few seconds.

So how does PayPal come into this? Most of these companies only accept PayPal. And PayPal blocks all accounts if it realises the IP is from Iran. Regardless of the amount, who is the receiver, how long the account has been used or behaviour of the account. Why? That is a a very good question, but maybe because someone is trying to purchase nuclear equipments or funding terrorism or ...! Honestly is that not silly? Paying 3$ for anti-censorship filter is illegal because you are connected from Iran? If anyone wants to use their PayPal account for illegal activities, they sure will use a proxy first so that they mask their IP. This will only affect ordinary people like me that are trying to pay for proxies as surely with the sanctions, you cannot buy anything that ships to Iran.

Now out of everyone, I am among the people least would wish to help Iran's government. My family was struck by this very government when my uncle working as Political Analyst for British Embassy was arrested by the authorities charged with spying back in 2009. He was then released after months but due to constant pressure and persecution from the authorities he had to flee from Iran and now has resumed his work in the Foreign Office in UK.

So where does this leave me?

Well it leaves my account blocked and closed. Having come back, I cannot use my account anymore. Emails I have sent have been responded with utter disinterent and "We don't care" attitude. And I have really hard time believing whether PayPal CEO does read complaint emails. I think I might have been rash with my tone in some emails but my frustration was extreme because of the unjustice.

But I am only one in many. Lives of many millions of Iranians have been affected by sanctions. Ordinary people suffer from the hands of the brutal government yet they find no consolation by the way they are treated outside Iran especially PayPal. For them, internet is the only way out of the oppression but blocking purchase of anti-censorship accounts is standing side-by-side with the Iranian regime. Does this make you happier Mr. David Marcus?

Take your Web API service consumption up to 11 with CacheCow.Client

ASP.NET Web API is here and a lot of teams have already started building software using it. If you have followed this and other blogs on webapibloggers.com, you probably have seen many various possibilities and avenues this framework brings for designing and building clean and scalable services.

ASP.NET Web API exposes all the goodness of HTTP. Caching is an important feature of HTTP and ASP.NET Web API allows for building services and clients that take advantage of this feature. I, along with a few friends in the Web API community, have been busy building caching extensions for Web API in a project called CacheCow which is hosted on GitHub.

CacheCow has two separate components: server and client. These will be used independently by service providers (server) and their consumers (client).

Server component allows for easy handling of HTTP caching scenarios on server by generating ETag, responding to validation of cache (see earlier posts on this subject, especially this and for full list here), cache invalidation and storage of cache metadata. Storage of cache metadata is possible in various stores, currently in-memory and SQL Server have been implemented and RavenDB and Redis is on the pipeline [UPDATE: RavenDB is implemented and NuGet package is available here]. Since storage has been abstracted away, any storage mechanism can be plugged in without making any server changes.

Client component looks after making cache-aware requests, cache validation and cache storage. Currently in-memory and file-based storage is available but other stores such as Redis, SQL Server, MongoDB and RavenDB are in the pipeline. Since storage has been abstracted away, any storage mechanism can be plugged in without making any client changes. One of important features of storage is total and per-site quota.

It is important to note that while clients can be browsers or native Apps (WPF, Silverlight, iOS, Android, etc), arguably more often than not they will be server components themselves. For example, an ASP.NET web site can call services of an ASP.NET Web API server. Also middleware components could similarly use resources exposed by Web API. As such it is very crucial that cache storage solutions are performant, scalable and configurable.

In this post, I will look into CacheCow.Client a little but more. For more info, you can read previous posts on the topic in this blog.

CacheCow.Client alternatives

The only alternative to CacheCow.Client (that I am aware of) is using WinINET caching. Internet Explorer also uses this so the cache store will be the same. This is basically windows' HTTP request stack which has been exposed in .NET Framework since v 2.0 through WebRequest:

RequestCachePolicy policy = 
        new RequestCachePolicy( RequestCacheLevel.Default);
WebRequest request = WebRequest.Create(uri);
request.CachePolicy = policy;
WebResponse response = request.GetResponse();

As you can see, we can define a cache policy which will be applied to the request and according to the policy, Internet Explorer cache is used. Cache policy has a few possible values that are defined here. Notable values include:

• CacheOnly: retrieves the request only from cache
• BypassCache: does not use cache at all and goes straight to the server
• CacheIfAvailable: retrieves from local or intermediate cache if resource available otherwise retrieve from server
• Default: Similar to previous but current cache policy takes effect 

This same mechanism is now exposed in HttpClient but basically is built on the top of WebRequest. Henrik fully covers this feature in his blog here.

Basically in order to use WinINET caching with the new Web API stack, you need to create an HttpClient but provide WebRequestHandler as the MessageHandler:

HttpClient client = new HttpClient(new WebRequestHandler()
                           {
                               CachePolicy = new RequestCachePolicy( RequestCacheLevel.Default)
                           });
// this is a sample. It is not advised to use .Result since can lead to deadlock!
var httpResponseMessage = client.GetAsync("http://carmanager.softxnet.co.uk/api/car/3").Result;
var httpResponseMessage2 = client.GetAsync("http://carmanager.softxnet.co.uk/api/car/3").Result;

Using this feature, you can enable caching with little coding on the client.

Why I would choose CacheCow.Client rather than WinINET

Because it goes to 11! As we saw, it is very easy to get started with caching in HttpClient. But as we noted, it is very likely that HttpClient could be used in a server context hence having a reliable and scalable solution is very important in production.

Here are a few advantages of CacheCow.Client over WinINET (or rather disadvantages of WinINET):

1. Caching will be shared with Internet Explorer

In a production scenario, you need an implementation which is predictable and reliable. If someone uses Internet Explorer on the machine, storage area for your application's resources will be taken by just simple browsing. This can lead Internet Explorer to flush application resources in order to store resources for the browsing session. 

2. You have little control over quota

With CacheCow.Client, you can define a global and a per-site quota for storage of resources while such feature is not accessible (although there could be some registry entries for changing these variables) in WinINET caching. Also these variables could be overwritten by installation of a newer version of Internet Explorer.

3. Cache is local to the machine and cannot be shared across servers

In a production scenario, it is desirable to be able to store caches in a central store so network traffic and requests could be limited while with WinINET caching, each server will use its own local cache store.

4. WinINET is file-based

With WinINET, cache is stored in a file location while for a high-throughput production environment, robust caching using solutions such as Redis is required. CacheCow client by abstracting the storage can use any number of storage mechanisms such as Redis, MongoDB, RavenDB, etc.

5. CachePolicy is global for the HttpClient instance

Sometimes you might need to bypass caching. With WinINET, this has to be done with changing policy at the client level which applies across all requests for that HttpClient while CacheCow.Client respects will not use cached resources if you set CacheControl header of the request to no-cache. This basically recommended implementation based on HTTP specification (RFC2616).

6. With WinINET you do not know if request was retrieved from cache

With WinINET, there is no way to tell if response was retrieved from the cache or origin server. CacheCow.Client provides x-cachecow header which provides various information which can be used for debugging and troubleshooting scenarios.

Introducing CacheCow.Client.FileCacheStore

Last week I finished first version of a persistent cache store which is file based. This is available using NuGet and (package name is CacheCow.Client.FileCacheStore) and the code available at GitHub.

Using this persistent store is very easy. After getting the package from NuGet, create an HttpCient while as a delegating handler, pass CachingHandler (covered before here) while setting the store to a new instance of FileStore. While creating a FileStore, you need to specify a folder for storing the cached resources:

var httpClient = new HttpClient(
 new CachingHandler(
 new FileStore("c:\\Cache"))
{
 InnerHandler = new HttpClientHandler()
});

That is all you have to do! Now all your requests will store cacheable resources in a file-based persistent store. 

Currently for quota it uses default values but I am in the process of exposing values so you can configure quota.

CacheCow roadmap

After exposing quota settings, I will be working on CacheCow.Client.RedisCacheStore for a high throughput production level cache storage.

Please keep me posted by your comments, feedback and raising bug/issues on the GitHub page. You are awesome!

Server-side Async: Careful with that Axe, Eugene

[Level T3]

In a previous post, I talked about the dangers lurking in doing server-side async operations in .NET 4.0. As you know, .NET 4.5 provides a much better syntax allowing async/await keywords to take your TPL Task-Soups to a much more readable and organised code. But even so, async will make debugging your application more difficult and bugs could take much longer to be reproduced, isolated and fixed.

Task-Soup

In .NET 4.0, when we add up continuations to create a chained task, we could end up with a few problems:

1. We could end up with an unobserved exception problem. This is nicely described by Ayende here
2. Nested lambda expressions could create unexpected problems with closure of variables
3. The code becomes hard to read.

On the third note, I will just bring an example from my own code in CacheCow. What is it that we are actually returning here?

return response.Then(r =>
{
 if (r.Content != null)
 {
  TraceWriter.WriteLine("SerializeAsync - before load",
   TraceLevel.Verbose);

  return r.Content.LoadIntoBufferAsync()
   .Then(() =>
   {
    TraceWriter.WriteLine("SerializeAsync - after load", TraceLevel.Verbose);
    var httpMessageContent = new HttpMessageContent(r);
    // All in-memory and CPU-bound so no need to async
    return httpMessageContent.ReadAsByteArrayAsync();
   })
   .Then( buffer =>
      {
       TraceWriter.WriteLine("SerializeAsync - after ReadAsByteArrayAsync", TraceLevel.Verbose);
       return Task.Factory.FromAsync(stream.BeginWrite, stream.EndWrite,
        buffer, 0, buffer.Length, null, TaskCreationOptions.AttachedToParent);                                                        
      }
     );

   ;
 }

Even looking at brackets gives me headache.

Is Async worth it at all?

Now we talk a lot about Async operations and its role in improving scalability. But really, is it worth it? How much scalability would it bring? Would it help or hinder?

The answer to these questions is yes, it does help. The more IO you do on your server-side actions, the more you benefit from improvement from scalability. So it is highly advisable to implement your ApiController actions as Async by returning Task or Task<T>. 

The truth is, it will help even with your non-IO-bound operations although it is not advisable to use Async in such scenarios. You can test it for yourself, create a sync and an async controller to do exactly the same operation and use a benchmarking tool to compare the performance.

I have a CarManager sample on GitHub which I use for testing CacheCow.Server and it contains two simple  controllers: CarController and CarAsyncController. All these do is to use an in-memory repository and their GET only looking up the dictionary by its key:

// sync version
public Car Get(int id)
{
 return _carRepository.Get(id);
}


// async version (on another controller)
public Task<Car> GetAsync(int id)
{
 return Task.Factory.StartNew(() => _carRepository.Get(id));
}

So if you use a benchmarking tool such as Apache Benchamrk ab.exe, you could see a slight increase in throughput using the async controller. In my case, there was a 10% increase in throughput using async.

My ordeal with a bug

Development of CacheCow has been marred by existent of a problem which as we will see, turns out to be not in my code. I have been battling with this for a few weeks (on and off) and could not progress CacheCow development because of that.

OK, here is how my story begins; I think the Sherlock Holmes nature of this troubleshooting could be amusing for others too. After realising that using simple ContinueWith will not flow the context (see previous post) I was tasked with changing all such cases with Then in the TaskHelpers which checks existence of SynchronizationContext and flows the context if it exists.

On the other hand, lostdev, one of CacheCow's most loyal users, informed me of an occasional null reference exception in CacheCow.Server. Now, I had already fixed a bug related to null reference exception when the a resource was being retrieved for the first time. I attributed the problem to the fix I had made and reported that the problem is fixed in the current version.

So I started developing file-based cache storage for CacheCow.Client (which will have its own post very soon) and replaced all ContinueWith cases with Then.

And then I started to experience deadlocks in CacheCow.Client when I was using file-based caching and sending concurrent GET requests to the server. As soon as I would remove FileStore, and replace with InMemoryCacheStore, it would work. So I started searching through the client code, debug, look at the threads, debug again, change code, debug... to no avail. As soon as I was using file-based caching it would start to appear so it had to be on the client.

Then I noticed a strange thing: I could only run 4 concurrent calls and rest would be blocked. Why? Then I started playing with the maxconnection property of the system.net configuration:

  <system.net>
 <connectionManagement>
   <add address = "*" maxconnection = "N" />
 </connectionManagement>
  </system.net>

and interestingly, by setting the N to a high number, I would get more concurrent connections - but only up to the number defined. Hmmm... so the requests do not quite finish. OK, I fired up Sysinternals' TcpView but unfortunately these connections did not show up (and I do not know why).

I was getting nowhere until I accidentally loaded an earlier version of the server code. To my surprise, I did not get the deadlock but this error which @Tugberk separately reported earlier but attributed to order of handlers:

[NullReferenceException: Object reference not set to an instance of an object.]
System.Web.Http.WebHost.HttpControllerHandler.EndProcessRequest(IAsyncResult result) +112
System.Web.Http.WebHost.HttpControllerHandler.System.Web.IHttpAsyncHandler.EndProcessRequest(IAsyncResult result) +10
System.Web.CallHandlerExecutionStep.OnAsyncHandlerCompletion(IAsyncResult ar) +129

OK, so it is probably happening on the server but the continuation code gets deadlocked on unhandled exception. I am close! So it was time to go to bed and I was positive that I would nail it the day after.

It was funny that I woke up the day after and with my in-bed reading on tweets, stumbled on @Tugberk's tweet on issue he had just created. That sounds exceedingly similar, so we just doubled checked our scenarios and it turned out that an HttpResponseMessage with empty RequestMessage property is not handled in Web API and a null reference exception is thrown at the end of the response clean-up code. And the reason I was seeing it only with file-based cache store was that the part of server-side code to return such responses was being triggered only using file-based store (since it was capable of persisting caches and was trying to validate the cache).

So as you can see, a seemingly unrelated problem can really confuse the nature of the bugs in async scenarios.

Conclusion

First of all, always use request.CreateResponse() instead of using new HttpResponseMessage. I googled for cases of new HttpResponseMessage and found +3000 entries. This is really dangerous and I think this is a bug in Web API and needs to be fixed. If you are using new, make sure you set the RequestMessage property.

And in general, be careful with doing server-side async operations. It is really a powerful axe but with it you are not quite sure what a slightly off swing could bring. Careful with that axe Eugene.

Going away

There are feelings that are so complex they don't have a name - unless we make one. They can be described, but not easily. Now "Going away" is one of those feelings.

Have you ever been to a place, on a holiday, and enjoying the places you visit and yet you cannot stop thinking about the fact that this is probably the last time you see this place? That is why you start taking pictures, videos, you want to register everything since you know your memory won't. It is likely that you would not go back to those pictures but having them is like you "own" those memories.

* * *

Beaumaris (map) is a tiny town in the remotest corner of the north west of Wales - in the Anglesey island. Once visiting Wales (we love Wales), on our hunt for nice places to eat, we found ourselves in there. Passing sailing boats on a green road with the Irish Sea on the right and stony walls on the left, we approached a town boasting a rich history - later found out about its Castle. We had no idea what we will find there.

Memory of that night was not captured on any picture. It did not need to. It felt like a surreal experience of two invisible tourists/observers entering an almost dream-like banquet. It was almost fictional. It reminds me of the opening (and closing) scene of the Russian Ark.

We parked in the corner of the square overlooking the Irish Sea and found our way through a small entrance to the main street. Ye Olde Bull's Head Inn was easy to find: a very old traditional pub with a classy hotel on the top. We ordered drinks and sat down in very old wooden chair and benches. While walking we were careful of our head: like all old pubs ceiling is low and wooden bars are visible.

And there was a banquet: a reunion of army forces of some sort - many of them American. Now out of  all places, why Beaumaris? Had they travelled all the way from the ocean to meet at such a remote place? I had no answer. I did chat to a few officers walking around there but my inquiry did not produce much more information. I have not been able to find any reference to this reunion on internet.

After 40 minutes, we were ushered to the brasserie where food was served. Suddenly atmosphere changed. It was a modern extension to the old building and an ethereal ambient light had made the chic contemporary design even more magical. Starter, food and dessert were excellent. Probably among best foods we ever had.

On leaving, I felt like "going away". But it did not matter. I had absorbed all those beautiful yet surreal moments.

* * *

I am visiting my family in Iran now. I do go back almost every year although now most my friends and my brother and sister are abroad.

My mother-in-law is ill. Medically speaking she is terminal. But everything is possible. Isn't it?

This is a chance to spend time with her and my parents. Things are not easy. Seeing suffering and not being able to cure is not easy - although there are things we can do. My wife is doing most of them so really there is not much for me to do. I occasionally help them out and I am just there, in case. My mother-in-law is a brave woman. She has been fighting and keeps fighting. This is so important for someone battling with a fierce illness. My wife is also very brave. She is fighting her emotions trying to help medically as much as she can - she is a doctor, a real one not like me who chickened out to do what he likes to do.

My holiday is running out - and the feeling of "going away" is coming back.

* * *

We visited Beaumaris again only after two years. This time there were no banquets but the place had no less magic: it was exactly as we had seen in that evening two years ago. Ye Olde pub with its crisp beers and the food excellent like before. 

I felt really special. It felt like an honour to visit the place again. I was fortunate enough to visit a "gone away" place.

And with my mother-in-law, I think it is possible - and perhaps definite. In this world. Or in the next.

Server-side TPL Async: Don't risk learning these lessons the hard way

[Level T2]

There has been more than a few times that I have felt I know all about TPL. Only to realise sometime later I was wrong, very wrong. Now you might read this and say to yourself "Come on, this is basic stuff. I know it well, thank you vety much". Well, it is possible that you could be right but I advise you carry on reading; what follows can surprise you.

None of what I am gonna talk about is new or it is being blogged about for the first time. Brad Wilson has an excellent series on the topic here but this is to serve as a digest of his posts targeted at a broader audience in addition to a few other points.

While this post is not directly related to ASP.NET Web API, most examples (and cases) are related to day-to-day scenarios we encounter in ASP.NET Web API.

Remember this covers pre-async/await keywords in .NET 4.5 and what you need to do if you are using .NET 4.0 and not async/await. Using async/await will cover you for some of the problems described below but not all.

Don't fire and forget

Tasks are ideal for decoupling pieces of functionality. For example, I can perform a database operation and at the same time audit the operation by outputing a log entry, writing to a file, etc. Using tasks I can de-couple these operations so that my database task returns without having to wait for audit to finish. This makes sense since database operation is high priority but audit is low priority:

private void DoDbStuff()
{
   CallDatabase();
   // doing audit entry asynchronously not to bog down database operation
   Task.Factory.StartNew(()=> AuditEntry("Database stuff was done"));
}

In fact, let's say we do not even care if audit is successful or not so we just fire and forget, it most audit will fail which is low priority. OK, it all seems innocent?

No! This innocent operation can bring down your application. Reason for it is that all async exceptions must be observed even if you do not care about them. If you don't, they will haunt you when you least expect them, at the time finalizer for task is run by GC. Such an unhandled exception will kill your app.

The link above talks about various ways of observing an exception. The most practical is to use a continuation and access the .Exception property of the task (just accessing the property is enough, does not need to do anything with the exception itself).

private void DoDbStuff()
{
   CallDatabase();
   // doing audit entry asynchronously not to bog down database operation
   Task.Factory.StartNew(()=> AuditEntry("Database stuff was done"))
      .ContinueWith(t => t.Exception); // fire and forget!
}

Another option which is more of a safe-guard against accidental unobserved exception, is to register to UnobservedTaskException on TaskScheduler:

 TaskScheduler.UnobservedTaskException +=
  (e, sender) => LogException(e);

So we register a handler to handle unobserved exceptions and this way they will be "observed". If you need to read more on this, have a look at Jon Skeet's post here.

This problem has made Ayende Rahien to run for the hills.

Respect SynchronizationContext

Uncle Jeffrey Richter tells us that

By default, the CLR automatically causes the first thread's execution context to flow to any helper threads.

And then we also learn that we can use ExecutionContext.SuppressFlow() to suppress flow of the thread context.

Now, what happens when we use ContinueWith()? It turns out unlike standard thread switches, context does not flow (I do not have a reference, if you do please let me know). This will help with improving performance of asynchronous task as we know context switching is expensive (and big part of it is context flow).

So why is it important? It is important because so many developers are used to HttpContext.Current. This context is stored in the thread storage area and passed along at the time of context switching. So if the context does not flow, HttpContext.Current will be null.

SynchronizationContext is a similar (but not same) concept. It is about a state that can be shared and used by different threads at the time of switching. I cannot explain this better than Stephen here. So using Post on SynchronizationContext ensures that the execution of continuation will happen in the same context and not necessarily by the same thread.

So basically the idea is that if you are in a Task pipeline (best example being MessageHandlers in ASP.NET Web API), you need to take responsibility for passing the context along the pipeline.

This is a snippet from ASP.NET Web API Source code that displays the steps. First of all you check to see if current context is null, if it is not then you have to use Post() to flow the context:

SynchronizationContext syncContext = SynchronizationContext.Current;

    TaskCompletionSource<Task<TOuterResult>> tcs = new TaskCompletionSource<Task<TOuterResult>>();

    task.ContinueWith(innerTask =>
    {
        if (innerTask.IsFaulted)
        {
            tcs.TrySetException(innerTask.Exception.InnerExceptions);
        }
        else if (innerTask.IsCanceled || cancellationToken.IsCancellationRequested)
        {
            tcs.TrySetCanceled();
        }
        else
        {
            if (syncContext != null)
            {
                syncContext.Post(state =>
                {
                    try
                    {
                        tcs.TrySetResult(continuation(task));
                    }
                    catch (Exception ex)
                    {
                        tcs.TrySetException(ex);
                    }
                }, state: null);
            }
            else
            {
                tcs.TrySetResult(continuation(task));
            }
        }
    }, runSynchronously ? TaskContinuationOptions.ExecuteSynchronously : TaskContinuationOptions.None);

    return tcs.Task.FastUnwrap();

There is a horrifying fact here. Most of the DelegatingHandler code out there (including some of mine) in various samples around internet do not respect this. Of course, looking at ASP.NET Web API source code reveals that they do indeed take care of this in their TaskHelper implementations and Brad tried to make us aware of it in his blog series. But I think we have not taken enough attention of the implications of ignoring SynchronizationContext.

Now my suggestion is to use the TaskHelpers and its extensions in the ASP.NET Web API (it is open source) or use the one provided in Brad's post. In any case,

Don't use Task for CPU-bound operations

Overhead of asynchronous operations is not negligible. You should only use async if you are doing an IO-bound operation (calling another web service/API, reading a file, reading a lot of data from database or running a slow query). I personally think even for normal IO operations, sync is more performant and scalable.

As we have talked about it here, the point about asynchronous programming on server-side is releasing the thread to be able to serve another request. Tasks are normally served by the CLR thread pool. If server already needs managed threads for its operations, it will be using CLR thread pool too. This means that by doing async operations you could be stealing threads needed for server's normal operations. A classic example is ASP.NET, so you should be careful to use async only if needed.

ContinueWith is Evil!

I think by now you should know why standard ContinueWith can be evil. First of all, it does not flow the context. Also it makes it easy for unboserved exceptions to creep into your code. My suggestion is to use .Then() from ASP.NET Web API's TaskHelpers.

Performance comparison

I think it is still early days - but I must say I would love to do a benchmark to quantify overhead of server-side asynchronous programming. Well if I do, this place will be where the result will first appear :)

So. Do I think I know all about TPL now? Hardly!