I am wondering what would be the best practice for deploying updates to a (MVC) Go web application. Imagine the following scenario :
1) Code and test some changes for my Go Web Application
2) Deploy update without anyone currently using the previous version getting interrupted.
I don't know how to make sure point 2) can be covered - when somebody is sending a request to the server and I rebuild/restart it just in this moment, he gets an error - even if the request just uses a part of the code I did not touch or that is backwards-compatible, or if I just added a new Request-handler.
Maybe I'm missing something trivial or a well-known pattern as I am just in the process of learning go and my previous web applications were ASP.NET- or php-applications where this was no issue as I did not need to restart the webserver on code changes.
It's not just an issue with Go, but in general we can divide the problem into two separate ones:
Making sure current requests do not get terminated and affect user experience.
Making sure there is no down-time in which new requests cannot be handled.
The first one is easier to tackle: You just don't violently kill your server, but tell it to exit, causing a "Drain phase", in which it does not accept new requests and only finishes the currently running requests, and exits. This can be done by listening on signals for example, and entering the app into a special state.
It's not trivial with Go as the default http server doesn't support shutting it down, but you can start a server with a net.Listener, and then keep a reference to it an close it when the time is due.
Now, doing only approach one and then starting the service again will cause new requests not to be accepted while this is going on, and we all know this can take a number of seconds in extreme cases.
So what we need is another instance of the server already running with the new code, the instant the old one is not responding to new requests, right? That can be done in several ways:
Having more than one server, and a load-balancer on top of them, allowing one (or more) server to take the load while we restart another. That's the simplest way, and the way most people do it. If you need N servers to take the load of your users, just keep N+1 and restart one at a time.
Using socket sharing tricks. In Newer Linux kernels, Many processes can listen and accept on the same port. What you do is simply start the new instance and then tell the old one to finish and exit. This way there is no pause. This is done by setting SO_REUSEPORT on the listening socket.
The above can be automated with ready to ship solutions, like Einhorn, that deals with all the details for you, see https://github.com/stripe/einhorn
Another approach is documented in this blog post: http://blog.nella.org/?p=879
Related
I have an ajax application built in catalyst running through fastcgi. I am noticing that all of my ajax requests spin up another process and that process sticks around indefinitely.
Ideally I would like to have my main page stick around but the ajax pages should just stop and restart as needed.
Is this a common issue and how does one get around this.
How do I tell the server to shut down inactive processes quicker?
The answer is web server dependent (and fastcgi module dependent for apache - as there are at least 2 different fcgi modules). You haven't told us which web server etc, so I can't really answer. (But consult the documentation for your web server / fastcgi module)
An alternative approach would of course be to run the server with external fastcgi, rather than in dynamic mode, which would mean you start a fixed number of processes up-front, but then never create more or less than this number to serve requests..
A web application I am developing needs to perform tasks that are too long to be executed during the http request/response cycle. Typically, the user will perform the request, the server will take this request and, among other things, run some scripts to generate data (for example, render images with povray).
Of course, these tasks can take a long time, so the server should not hang for the scripts to complete execution before sending the response to the client. I therefore need to perform the execution of the scripts async, and give the client a "the resource is here, but not ready" and probably tell it a ajax endpoint to poll, so it can retrieve and display the resource when ready.
Now, my question is not relative to the design (although I would very much enjoy any hints on this regard as well). My question is: does a system to solve this issue already exists, so I do not reinvent the square wheel ? If I had to, I would use a process queue manager to submit the task and put a HTTP endpoint to shoot out the status, something like "pending", "aborted", "completed" to the ajax client, but if something similar already exists specifically for this task, I would mostly enjoy it.
I am working in python+django.
Edit: Please note that the main issue here is not how the server and the client must negotiate and exchange information about the status of the task.
The issue is how the server handles the submission and enqueue of very long tasks. In other words, I need a better system than having my server submit scripts on LSF. Not that it would not work, but I think it's a bit too much...
Edit 2: I added a bounty to see if I can get some other answer. I checked pyprocessing, but I cannot perform submission of a job and reconnect to the queue at a later stage.
You should avoid re-inventing the wheel here.
Check out gearman. It has libraries in a lot of languages (including python) and is fairly popular. Not sure if anyone has any out of the box ways to easily connect up django to gearman and ajax calls, but it shouldn't be do complicated to do that part yourself.
The basic idea is that you run the gearman job server (or multiple job servers), have your web request queue up a job (like 'resize_photo') with some arguments (like '{photo_id: 1234}'). You queue this as a background task. You get a handle back. Your ajax request is then going to poll on that handle value until it's marked as complete.
Then you have a worker (or probably many) that is a separate python process connect up to this job server and registers itself for 'resize_photo' jobs, does the work and then marks it as complete.
I also found this blog post that does a pretty good job summarizing it's usage.
You can try two approachs:
To call webserver every n interval and inform a job id; server processes and return some information about current execution of that task
To implement a long running page, sending data every n interval; for client, that HTTP request will "always" be "loading" and it needs to collect new information every time a new data piece is received.
About second option, you can to learn more by reading about Comet; Using ASP.NET, you can do something similiar by implementing System.Web.IHttpAsyncHandler interface.
I don't know of a system that does it, but it would be fairly easy to implement one's own system:
create a database table with jobid, jobparameters, jobresult
jobresult is a string that will hold a pickle of the result
jobparameters is a pickled list of input arguments
when the server starts working on a job, it creates a new row in the table, and spwans a new process to handle that, passing that process the jobid
the task handler process updates the jobresult in the table when it has finished
a webpage (xmlrpc or whatever you are using) contains a method 'getResult(jobid)' that will check the table for a jobresult
if it finds a result, it returns the result, and deletes the row from the table
otherwise it returns an empty list, or None, or your preferred return value to signal that the job is not finished yet
There are a few edge-cases to take care of so an existing framework would clearly be better as you say.
At first You need some separate "worker" service, which will be started separately at powerup and communicated with http-request handlers via some local IPC like UNIX-socket(fast) or database(simple).
During handling request cgi ask from worker state or other data and replay to client.
You can signal that a resource is being "worked on" by replying with a 202 HTTP code: the Client side will have to retry later to get the completed resource. Depending on the case, you might have to issue a "request id" in order to match a request with a response.
Alternatively, you could have a look at existing COMET libraries which might fill your needs more "out of the box". I am not sure if there are any that match your current Django design though.
Probably not a great answer for the python/django solution you are working with, but we use Microsoft Message Queue for things just like this. It basically runs like this
Website updates a database row somewhere with a "Processing" status
Website sends a message to the MSMQ (this is a non blocking call so it returns control back to the website right away)
Windows service (could be any program really) is "watching" the MSMQ and gets the message
Windows service updates the database row with a "Finished" status.
That's the gist of it anyways. It's been quite reliable for us and really straight forward to scale and manage.
-al
Another good option for python and django is Celery.
And if you think that Celery is too heavy for your needs then you might want to look at simple distributed taskqueue.
I'm trying to create a small and basic "ajax" based multiplayer game. Coordinates of objects are being given by a PHP "handler". This handler.php file is being polled every 200MS, by using ajax.
Since there is no need to poll when nothing happens, I wonder, is there something that could do the same thing without frequent polling? Eg. Comet, though I heard that you need to configure server side applications for Comet. It's a shared webserver, so I can't do that.
Maybe prevent the handler.php file from even returning a response if nothing has to be changed at the client, is that possible? Then again you'd still have the client uselessly asking for a response even though something hasn't changed yet. Basically, it should only use bandwidth and sever resources if something needs to be told to the client, eg. the change of an object's coordinates.
Comet is generally used for this kind of thing, and it can be a fragile setup as it's not a particularly common technology so it can be easy not to "get it right." That said, there are more resources available now than when I last tried it ~2 years ago.
I don't think you can do what you're thinking and have handler.php simply not return anything and stop execution: The web server will keep the connection open and prevent any further polling until handler.php does something (terminates or provides output). When it does, you're still handling a response.
You can try a long polling technique, where your AJAX allows a very large timeout (e.g. 30 seconds), and handler.php spins without responding until it has something to report, then returns. (You'll want to make sure the spinning is not resource-intensive). If handler.php "expires" and nothing happens, have it exit and let AJAX poll again. Since it only happens every 30 seconds, it will be a huge improvement over ~5 times a second. That would keep your polling to a minimum.
But that's the sort of thing Comet is designed for.
As Ajax only offers you a client server request model (normally termed pull, rather than push), the only way to get data from the server is via requests. However a common technique to get around this is for the server to only respond when it has new data. So the client makes a request, the server hangs on to that request until something happens and then replies. This gets around the need for frequent polling even when the data hasn't changed as you only need the client send a new request after it gets a response.
Since you are using PHP, one simple method might be to have the PHP code call the sleep command for 200ms at a time between checks for data changes and then return the data to the client when it does change.
EDIT: I would also recommend having a timeout on the request. So if nothing happens for say 2 seconds, a "no change" message is sent back. That way the client knows the server is still alive and processing its request.
Since this is tagged “html5”: HTML5 has <eventsource> and WebSocket, but the implementation side is still in the future tense in practice.
Opera implemented an old version of <eventsource> called <event-source>.
Here's a solution - use a SaaS comet provider, such as WebSync On-Demand. No server resources to worry about, shared hosting or not, since it's all offloaded, and you can push out the information as needed.
Since it's SaaS, it'll work with any server language. For PHP, there's already a publisher written and ready to go.
The server must take part in this. Check with the hosting provider what modules are available. Or try to convince them to support Comet.
Maybe you should consider a small Virtual Private Server (VPS) for this.
One thing to add on the long polling suggestions: If you're on a shared server, this solution will have limited scalability, as each active long poll will keep a connection (and a server-side process to service that connection) active. Your provider most likely has limits (either policy-defined or de facto) on the number of connections you can have open at a time, so you'll hit a wall if you have more sessions/windows than that playing concurrently.
I'm working on a consumer web app that needs to do a long running background process that is tied to each customer request. By long running, I mean anywhere between 1 and 3 minutes.
Here is an example flow. The object/widget doesn't really matter.
Customer comes to the site and specifies object/widget they are looking for.
We search/clean/filter for widgets matching some initial criteria. <-- long running process
Customer further configures more detail about the widget they are looking for.
When the long running process is complete the customer is able to complete the last few steps before conversion.
Steps 3 and 4 aren't really important. I just mention them because we can buy some time while we are doing the long running process.
The environment we are working in is a LAMP stack-- currently using PHP. It doesn't seem like a good design to have the long running process take up an apache thread in mod_php (or fastcgi process). The apache layer of our app should be focused on serving up content and not data processing IMO.
A few questions:
Is our thinking right in that we should separate this "long running" part out of the apache/web app layer?
Is there a standard/typical way to break this out under Linux/Apache/MySQL/PHP (we're open to using a different language for the processing if appropriate)?
Any suggestions on how to go about breaking it out? E.g. do we create a deamon that churns through a FIFO queue?
Edit: Just to clarify, only about 1/4 of the long running process is database centric. We're working on optimizing that part. There is some work that we could potentially do, but we are limited in the amount we can do right now.
Thanks!
Consider providing the search results via AJAX from a web service instead of your application. Presumably you could offload this to another server and let you web application deal with the content as you desire.
Just curious: 1-3 minutes seems like a long time for a lookup query. Have you looked at indexes on the columns you are querying to improve the speed? Or do you need to do some algorithmic process -- perhaps you could perform some of this offline and prepopulate some common searches with hints?
As Jonnii suggested, you can start a child process to carry out background processing. However, this needs to be done with some care:
Make sure that any parameters passed through are escaped correctly
Ensure that more than one copy of the process does not run at once
If several copies of the process run, there's nothing stopping a (not even malicious, just impatient) user from hitting reload on the page which kicks it off, eventually starting so many copies that the machine runs out of ram and grinds to a halt.
So you can use a subprocess, but do it carefully, in a controlled manner, and test it properly.
Another option is to have a daemon permanently running waiting for requests, which processes them and then records the results somewhere (perhaps in a database)
This is the poor man's solution:
exec ("/usr/bin/php long_running_process.php > /dev/null &");
Alternatively you could:
Insert a row into your database with details of the background request, which a daemon can then read and process.
Write a message to a message queue which a daemon then read and processed.
Here's some discussion on the Java version of this problem.
See java: what are the best techniques for communicating with a batch server
Two important things you might do:
Switch to Java and use JMS.
Read up on JMS but use another queue manager. Unix named pipes, for instance, might be an acceptable implementation.
Java servlets can do background processing. You could do something similar to this technology in a web technology with threading support. I don't know about PHP though.
Not a complete answer but I would think using AJAX and passing the 2nd step to something thats faster then PHP (C, C++, C#) then a PHP function pick the results off of some stack most likely just a database.
This is my setting: I have written a .NET application for local client machines, which implements a feature that could also be used on a webpage. To keep this example simple, assume that the client installs a software into which he can enter some data and gets some data back.
The idea is to create a webpage that holds a form into which the user enters the same data and gets the same results back as above. Due to the company's available web servers, the first idea was to create a mono webservice, but this was dismissed for reasons unknown. The "service" is not to be run as a webservice, but should be called by a PHP script. This is currently realized by calling the mono application via shell_exec from PHP.
So now I am stuck with a mono port of my application, which works fine, but takes way too long to execute. I have already stripped out all unnecessary dlls, methods etc, but calling the application via the command line - submitting the desired data via commandline parameters - takes approximately 700ms. We expect about 10 hits per second, so this could only work when setting up a lot of servers for this task.
I assume the 700m are related to the cost of starting the application every time, because it does not make much difference in terms of time if I handle the request only once or five hundred times (I take the original input, vary it slighty and do 500 iterations with "new" data every time. Starting from the second iteration, the processing time drops down to approximately 1ms per iteration)
My next idea was to setup the mono application as a remoting server, so that it only has to be started once and can then handle incoming requests. I therefore wrote another mono application that serves as the client. Calling the client, letting the client pass the data to the server and retrieving the result now takes 344ms. This is better, but still way slower than I would expect and want it to be.
I have then implemented a new project from scratch based on this blog post and get stuck with the same performance issues.
The question is: am I missing something related to the mono-projects that could improve the speed of the client/server? Although the idea of creating a webservice for this task was dismissed, would a webservice perform better under these circumstances (as I would not need the client application to call the service), although it is said that remoting is faster than webservices?
I could have made that clearer, but implementing a webservice is currently not an option (and please don't ask why, I didn't write the requirements ;))
Meanwhile I have checked that it's indeed the startup of the client, which takes most of the time in the remoting scenario.
I could imagine accessing the server via pipes from the command line, which would be perfectly suitable in my scenario. I guess this would be done using sockets?
You can try to use AOT to reduce the startup time. On .NET you would use ngen for that purpoise, on mono just do a mono --aot on all assemblies used by your application.
AOT'ed code is slower than JIT'ed code, but has the advantage of reducing startup time.
You can even try to AOT framework assemblies such as mscorlib and System.
I believe that remoting is not an ideal thing to use in this scenario. However your idea of having mono on server instead of starting it every time is indeed solid.
Did you consider using SOAP webservices over HTTP? This would also help you with your 'web page' scenario.
Even if it is a little to slow for you in my experience a custom RESTful services implementation would be easier to work with than remoting.