I am using a PersistentConnection for publishing large amounts of data (many small packages) to the connected clients.
It is basically a one way direction of data (since each client will call endpoints on other servers to set up various subscriptions, so they will not push any data back to the server via the SignalR connection).
Is there any way to detect that the client cannot keep up with the messages sent to it?
The example could be a mobile client on a poor connection (e.g. in a roaming situation, the speed may vary a lot). If we are sending 100 messages per second, but the client can only handle 10, we will eventually lose the messages (due to the message buffer on the server side).
I was looking for a server side event, similar to what has been done on the (SignalR) client, e.g.
protected override Task OnConnectionSlow(IRequest request, string connectionId) {}
but that is not part of the framework (for good reasons, I assume).
I have considered using the approach (suggested elsewhere on Stackoverflow), to let the client tell the server (e.g. every 10-30 seconds) how many messages it has received, and if that number differentiates a lot from the number of messages sent to the client, it is likely that the client cannot keep up.
The event would be used to tell the distributed backend that the client cannot keep up, and then turn down the data generation rate.
There's no way to to this right now other than coding something custom. We have discussed this in the past as a potential feature but it isn't anywhere the roadmap right now. It's also not clear what "slow" means as it's up to the application to decide. There'd probably be some kind of bandwidth/time/message based setting that would make this hypothetical event trigger.
If you want to hook in at a really low level, you could use owin middleware to replace the client's underlying stream with one that you owned so that you'd see all of the data going over the write (you'd have to do the same for websockets though and that might be non trivial).
Once you have that, you could write some time based logic that determined if the flush was taking too long and kill the client that way.
That's very fuzzy but it's basically a brain dump of how a feature like this could work.
Related
Not sure what category this question falls into; perhaps general networking / design / algorithms.
For a project I am looking at having one server with multiple connected clients. After some time, when all clients have connected, the server should send a message to each client instructing them to take some action. I need to guarantee that each client will execute this action at exactly the same time. Theoretically, how can this be done? What are the practical complications I will come up against? My target platform is mobile.
One solution I can think of;
The server actively and continuously keep track of the round-trip latency for each client. Provided this latency doesn't change too fast over time, the server should be able to compensate for each client's lag and send messages to each such that they all start execution at roughly the same time. Is there a better way?
One not-really related question: Client side and server side events not firing simultaneously
It can easily be done.
You don't care about latency nor you need the same machine time at clients.
The key here is to create a precise appointment.
Since clients communicate to the server, and not vice versa (you didn't say anything about it though). I can give you the following solution:
When a client connects to the server, it should send their local time.
When the server thinks it's time for the event to be set. It should send an appointment event to each client, with their local time in it. Server can calculate this.
Then, each client knows when exactly they need to do something by setting a timer till the time for their appointment comes.
In theory yes you can but not in real life.
At least you should add some a validity time-slot. All actions should be in that predefined time-slot in order that action to be valid.
So basically "same moment" = "a predefined time slot".
A predefined time-slot can be any value that is close to same moment or real-time.
I'm using SignalR with Redis as a message bus on a server that sits behind an Nginx proxy for load balancing. I used SignalR's PersistentConnection class to write a simple chat program that broadcasts messages to users belonging to the same certain group. Users are added to a group in OnConnectedAsync, removed in OnDisconnectAsync, and the user-to-group mapping is deterministic.
Currently, the client side falls back to long polling for whatever reason (I'm not entirely sure why), and whenever the client sets up a new connection after waiting for and receiving a response, seemingly at random, the server will sometimes respond to the new connection immediately with the previous response, despite there having only been one POST.
The message ID's tend to differ by exactly one, (the smaller ID coming first), with the rest of the response remaining the same. I logged some debug info and am quite positive that my override of OnReceivedAsync is sending one response per one request. I tried the same implementation without the Redis message bus, and got the same problem. Running locally (with long polling) however yielded good results so I suspect that the problem might be with the way the message bus might be buffering messages to refresh clients who might not be caught up, and some weird timing with the cutting/setting up of connections with the Nginx load balancer, but beyond that, I am very much at a loss.
Any help would be appreciated.
EDIT: Further investigation reveals that duplication occurs at somewhat regular intervals of approximately 20-30 seconds. I'm led to believe that the message expiration in the message bus might have something to do with the bug.
EDIT: Bug can be seen here: http://tinyurl.com/9q5t3va
The server is simply broadcasting a counter being sent by the client. You will notice some responses are duplicated every 20 or so.
Reducing the number of worker processes in the IIS (6.0) Server Manager from 2 to 1 solved the problem.
How Facebook, Google plus or other informations web site, constantly retrieves information from the stream?
I suppose there is an asynchronous recovery , but how he gets constantly? It's like an infinite loop?
Which technology is used ?
There are a few different approaches to displaying updates in near-real time on the web. Here are some of the most common ones:
Short polling
The simplest approach to the problem is to continuously poll the server on a short interval (hence the name). This means that every few seconds, client-side code sends an asynchronous request to the server and displays the result. The downside to this approach is that if updates happen less frequently than the server is queried, the client is doing a lot of work for little payoff. There may also be a slight delay between when the event happens on the server and when the client receives it, based on the polling frequency.
Long polling
The next evolutionary step from short polling is what's known as long polling, where the client-side JavaScript fires off an asynchronous request to the server as soon as the page loads. The server only responds to the request when an update is made, and once the response reaches the client, another request is fired off immediately. The key part of this approach is that the asynchronous request can wait for the server for a long time.
Long polling saves bandwidth and computation time, since the response is only handled when the server has something that changed. It does require more complex server-side logic, but it does allow for near-instant updates on the client side.
This question has a decent sample: How do I implement basic "Long Polling"?
WebSockets
WebSockets are a relatively new technology, and allow for two-way communication in a way that's similar to standard network sockets. The server or client can send messages across the socket that trigger events on the other side of the connection. As nice as this is, browser support isn't as widespread enough to make it a dependable solution.
For the current WebSocket specification, take a look at RFC 6455.
I have created a simple chat server that is driven by client polling. Clients send requests for data every few seconds, and get handed any new messages as well as information about whether their peer is still connected.
Since the client is running on a mobile platform (iPhone), I've been looking for ways of getting rid of the polling, which quickly drains the battery. I've read that it's possible to keep an http connection open indefinitely, but haven't understood how to utilize this technique in practice. I'm also wondering whether such connections are stable enough to use in a mobile setting.
The ideal scenario would be that the server only sends data to clients when an event that affects them has occurred (such as a peer posting a message or going off line).
Is it advisable to try to accomplish this over http, or would I have to write my own protocol over tcp? How hard would it be to customize xmpp to my need (my chat server has some specialized features that I would have to easily implement).
How about push technology? see http://en.wikipedia.org/wiki/Comet_(programming)
I think you're describing XMPP over BOSH.
http://xmpp.org/extensions/xep-0206.html
I've used this http-binding method between a chat server and javascript client on non-mobile devices. It worked well for me.
You might like to check out this project which uses a variety of techniques including Comet. Release details are here, here's a snippet from that page
It’s my distinct pleasure to be able
to announce the first public showing
of a project that I’ve been working on
in my spare time in the last month or
two, a new Web Based IRC chat
application.
This project brings together a lot of
new technologies which had to be
developed to make this a feasible,
scalable and efficient.
Some of the underlying tools build to
make this posible that i consider
’stable enough’ are already released,
such as the php Socket Daemon library
i wrote to be able to deal with
hundreds up to many thousands of
“Comet” http connections, and an equal
amount of IRC client connections.
I just found this article myself, which describes the following technique (which I referred to in the question):
... have the client make an HTTP request
and have the server hold the request
on the queue until there is a message
to push. if the TCP/IP connection is
lost or times-out, the client will
make a new HTTP request, and the delay
will only be the round trip time for a
request/response pair . . . this model
effectively requires two TCP/IP
connections for HTTP, client to
server, though none permanent and
hence mobile friendly
I think this is nearly impossible and dangerous. The internet works stateless and connectionless meaning that the connection between client and server is always handled as unreliable. And this is not for fun.
By trying to get a stateful connection you are introducing new issues. Especially from a 3g application. What if the connection breaks? You have no control over the server and cannot push.
I think it would even be easier to send sms/text messages and have an application that handles that.
What's a good way to connect the synchronous http request/response model with an asynchronous queue based model?
When the user's HTTP request comes it generates a work request that goes onto a queue (beanstalkd in this case). One of the workers picks up the request, does the work, and prepares a response.
The queue model is not request/response - there are only requests, not responses. So the question is, how best do we get the response back into the world of HTTP and back to the user?
Ideas:
Beanstalkd supports light weight topics or queues (they call them tubes). We could create a tube for each request, have the worker create a message on that tube, and have the http process sit and wait on the tube for the response. Don't particularly like this one since it has apache processes sitting around taking memory.
Have the http client poll for the response. The user's initial HTTP request kicks off the job on the queue and returns immediately. The client (the user's browser) polls periodically for a response. On the backend the worker puts its response into memcached, and we connect nginx to memcached so the polling is light weight.
Use Comet. Similar to the second option, but with fancier http communication to avoid polling.
I'm leaning towards 2 since it's easy and well know (I haven't used comet yet). I'm guessing there's probably also a much better obvious model I haven't thought of. What do you think?
Here's how to implement request-response efficiently on JMS which might be helpful (though Java/JMS centric). The general idea is to create a temporary queue per client/thread then use correlationIDs to correlate requests to replies etc.
Polling is the simple solution; comet is the more efficient solution. You've got it nailed :)
I personally love comet (although I'm biased, since I helped write WebSync), it nicely lets your clients subscribe to a channel and get the message when your server process is ready. Works like a champ.
I'm looking to implement a Beanstalkd and memcached system to run a number of processes following a request - in this case, looking up information when a user logs in (the number of messages a user has waiting for example). The info is stored in Memcached and then read back on the next page load.
Without knowing a little more about what tasks you are doing though, it's not so easy to say what needs to be done, or how. Option #2 is however the simplest, and that may be all you need - depending on what you are pushing back into the workers.