I'm currently using MediaFoundation to read MP4 video from a remote server using HTTP. I create the media source using the source resolver method CreateObjectFromURL.
This works perfectly fine and the video is great, except for one huge pain point:
All data received by the source seems to be cached to disk, specifically in ...AppData\Local\Microsoft\Windows\INetCache\IE. This is not ideal as the streams can be open for an unlimited amount of time because the video is live.
At the moment the only real solution is to periodically disconnect, clean up temporary internet files, then reconnect again. This would need to be done every few hours to stop the machine from filling up and exploding. It is not really an acceptable long term solution for the end user of course.
I have tried to disable caching by setting the MFNETSOURCE_CACHEENABLED attribute on the media source to false, but it doesn't seem to stop it caching at all.
Is there a trick I'm missing somewhere?
Related
I have this intriguing problem on Azure Website. My website uses 4 script files and 3 style files, each minified. They are not so big, bigest has near 200 KBs. Website had already started. Azure's Always On option is turned on. When I call to WebApi for data it returns in <50ms.
And when app is reloaded it needs 250 ms just to get first byte from tiniest script, and others needs much more. Initial Html is loaded in 60 ms. Scripts/styles are cached so they are not downloaded, but the TTFB time is killing the performance. This repeats every single reload. App is not containing any sophisticated configuration so it should run much faster than it.
What can cause such problems?
Although your static files are cached, the browser still issues requests with if-modifies-since header (which results in a 304).
While it doesn't need to download the actual content, it still needs to wait the RTT + server think time to continue.
I would suggest two things:
Adding Cache-Control and Expire headers - will help avoid 304 in some cases (pretty much unless you hit F5)
Using a proper CDN - such as Incapsula or others, that will minimize the RTT + think time. It can also be used to easily control cache settings for various resources.
More good stuff here.
Good Luck!
From here:
As you saw earlier, IIS 7 caches the compressed versions of static
files. So, if a request arrives for a static file whose compressed
version is already in the cache, it doesn’t need to be compressed
again.
But what if there is no compressed version in the cache? Will IIS 7
then compress the file right away and put it in the cache? The answer
is yes, but only if the file is being requested frequently. By not
compressing files that are only requested infrequently, IIS 7 saves
CPU usage and cache space.
By default, a file is considered to be requested frequently if it is
requested two or more times per 10 seconds.
So, the reason your users are being served an uncompressed version of the javascript file is because it didn't meet the default threshold for being compressed; in other words, the javascript file was not requested 2 times within 10 seconds.
To control this, there is one attribute we must change on the <serverRuntime> element, which controls compression: frequentHitThreshold. In order for your file to be compressed when it is requested once, change your <serverRuntime> element to look like this:
<serverRuntime enabled="true" frequentHitThreshold="1" />
This will slightly impact your CPU performance if you have many javascript files that are being served and you have users quite often, but likely if you have users often enough to impact CPU from compressing these files, then they are already compressed and cached!
My guess would be Azures always on.
If it works anything like the one CloudFlare provides, it essentially proxies the request and tries to cache it.
Depending on the exact implementation of this cache on the side of Azure, it might wait for the scripts output to complete to cache it/validate the cache and then pass it on to the browser.
You might have a chance checking the caching configuration and disable always on for your scripts if possible.
The scripts and styles are static files and by default are compressed (you can check this with HTTP header "content-encoding": gzip) before being sent to client. So, the TTFB consists of network latency, browser HTTP channel scheduling and the static file compression time from server.
On the other hand, your Web API data is dynamic data and by default is not compressed, so possible its TTFB is less than the TTFB for static files.
However, you don't need to switch off static compressing, otherwise TTFB is minimized but content transferring time will be extended. Actually, you don't need to worry about TTFB, see more info: https://blog.cloudflare.com/ttfb-time-to-first-byte-considered-meaningles/
I finished with storing files on Azure Storage and serving them by Azure CDN. It provides high speed of response and costs nothing. I add them to blob every publish, in Pre-build event by Gulp.
well... there are 2 main problems with your site:
you are using AZURE - a high priced service with a poor performance.... don't ask me why people think that this is a good service
you are storing client files side-by-side with the server files.. while server files should be stored in a specific server, client files can practically can be served from... everywhere
so - please use a CDN (or any other server) for your client side files (mainly css and js, you may consider moving fonts and images as well)
Abstract: There is a page with a player that loads audio file and plays it. The player used on the web page is jwplayer. I need to find a way to determine if the audio file is being streamed to the player or not.
Background: In my research I found that if I use nginx header like X-Accel-Redirect - the file will be streamed. I have setup the web server with nginx + apache combination (nginx is reverse proxy for apache), after that I pointed jwplayer to the mp3 file - and it is working. I mean I am able to click anywhere on the audio timeline and it immediately starts playing sound. But, since I didn't set that header yet, and adding the fact that player already works - that's why I need to check my question and know for sure.
Some of my own thoughts: JwPlayer itself supports some kind of bufferring, so I have no idea whether it just downloads the mp3 file I am testing this functions on, or it receives the stream and plays it out.
Is there a way to check and know for sure? The only idea about all of this I have is to check access logs, but I don't know what to look for, or if I need a special format for the logs to see those requried data.
While I was researching the issue I got some weird download related topics and something about HTTP headers with "Ranges" in them, but I am not sure that it relates to the streaming or not.
Please advice.
From the point of view of the server, there is no difference between download and streaming. A server just send bits. What happens to those bits later is unknown. What you need is a player that sends reports to back to the server or a loging service such as mixpanel.
I have a dynamically generated rss feed that is about 150M in size (don't ask)
The problem is that it keeps crapping out sporadically and there is no way to monitor it without downloading the entire feed to get a 200 status. Pingdom times out on it and returns a 'down' error.
So my question is, how do I check that this thing is up and running
What type of web server, and server side coding platform are you using (if any)? Is any of the content coming from a backend system/database to the web tier?
Are you sure the problem is not with the client code accessing the file? Most clients have timeouts and downloading large files over the internet can be a problem depending on how the server behaves. That is why file download utilities track progress and download in chunks.
It is also possible that other load on the web server or the number of users is impacting server. If you have little memory available and certain servers then it may not be able to server that size of file to many users. You should review how the server is sending the file and make sure it is chunking it up.
I would recommend that you do a HEAD request to check that the URL is accessible and that the server is responding at minimum. The next step might be to setup your download test inside or very close to the data center hosting the file to monitor further. This may reduce cost and is going to reduce interference.
Found an online tool that does what I needed
http://wasitup.com uses head requests so it doesn't time out waiting to download the whole 150MB file.
Thanks for the help BrianLy!
Looks like pingdom does not support the head request. I've put in a feature request, but who knows.
I hacked this capability into mon for now (mon is a nice compromise between paying someone else to monitor and doing everything yourself). I have switched entirely to https so I modified the https monitor to do it. The did it the dead-simple way: copied the https.monitor file, called it https.head.monitor. In the new monitor file I changed the line that says (you might also want to update the function name and the place where that's called):
get_https to head_https
Now in mon.cf you can call a head request:
monitor https.head.monitor -u /path/to/file
I want to at least prevent normal users to download my flash video.
What's the best way to do it?
Create a httphandler, add a token (e.g. timeid), set the cache control to no-cache so that only the users with correct token can view the correct video. Is that feasible?
It is the requirement from client that the video should not be downloaded by users and should be watched only in the particular website.
I want to know if this works:
http://www.somesite.com/video.swf?time=1248319067
Server will generate a token(time in the above example) so that user can only have one request to this link. If the user wants to watch the video again, he needs to go to our website to get the token again. Is this okay to prevent novices from downloading?
I can't download this flash video by the downloadHelper firefox plugin:
http://news.bbc.co.uk/2/hi/americas/8164177.stm
Updated (13:49 pm 2009/07/23):
The above file can be downloaded using some video download software.
The video files of following Chinese sites are well protected (I can't download it using many video download software):
http://programme.tvb.com/drama/abrideforaride/video/
Do you know how it is done?
I dont think there is an easy way to stop people from getting your videos if they want them,
there are plenty of plugins for firefox that allow downloading from even youtube and many places. And i imagine those plugins would disable any attempt you made to hide your videos.
not too terribly different than taking an image from flicker, they put a clear gif image over the image that you want to view, so that when you right click and save you get "the shield" image, however can be defeated by the lowly print screen button.
if you want casual users from getting your file, use a flash control and buffer a minute or two of your videos and make that flash authenticate with the server to get those files. that seems reasonable to me
I don't think there really is an easy way to limit people from getting at it. Your sending them the video, that is how they are able to view it. Any user could just use FRAPS or a similar tool to copy the video from the screen as well.
If your worry is being copied and used elsewhere then you can watermark it or use a few other types of copy protection methods that will allow you to identify your work on other sites. If your worried about people copying it for personal use, then you really have no way of stopping it, you are sending it to them.
Edit: Due diligence would be to inform your customer of how easy it is to copy the work that they will be posting. Most clients have really no idea how easy it is.
This is how I like to tackle this issue.
This method works by creating a ticket to download the content over one http request...Another attempt to use the same ticket to download the content will fail, hence any extensions that attempt to download the content again or a user manually attempting to fail to do so, hence the flash player will be the only way to download the content. However there is one downfall for this approach, users will not be able to skip to a part of the video that has not been download...in some standard player implementation that may even stop the video from loading. Any ideas on this will be highly appreciated.
I begin by writing a PHP script that takes in a video_id, file_name, or a local path to your video file (Depending on the storage infrastructure of your video collection) in a GET request along with a unique hash value (a hard to guess and come up with probably generated with a secret key so it can be validated to be coming from our reciever (flash player), if the hacker send us a used hash or an invalid hash (does not satisfy our key), we will not send him the file). The PHP script then opens the video file and sends its content with the correct video mime type. for FLV the mime type is video/x-flv. It makes sure that once a unique hash has not been used before and is validly generated from your secret encryption key.
Then once the page with the flash player is loading we can give the .php file with the right get parameters as the video url to the video player. (If it is a prude player that only allows flv files you can always program your .htaccess file to parse .flv files as php script in the specific folder only, and rename your .php file as .flv and try your luck)...anyways...Also generate a hash key...perhaps you can take the servers current time and append it to a salt value such as another key known by both scripts, and encrypt this final concatenation with your secret key.
So once the video gateway php script will recieve a filename or hash key...it will decrypt the hash key and figure out if it is validly generated from teh sister script, and make sure not to send the video again to the same hash key...
For added security you can perhaps reset the secret key everyday using either a cronjob or bootstrap mechanism. To prevent duplicate use of hashkeys you can store them in a mysql database, file operations, or NOSQL (depending on your needs and infrastructure).
Make sure that the file is requested by the same user agent the hash key was generated for. In case the hacker trys to cURL or Wget your videos unused url before the flash player gets a chance to consume the hash key. In this case the hacker will have to imitate the browser's user agent or download the file using their command line tool as well...However please note that this is not your average champ.
It sounds like you need to add authorization and authentication.
You could put the flash video under a different folder in your ASP.Net application and add a web.config file in that folder to deny access to unauthorized users. For example:
Then you need to enable authentication for your website. The simplest method is forms authentication. A trivial example with hard coded username and password is provided here.
There is loads that you can do with the authentication framework in ASP.Net I suggest googling a bit.
The only way to do this is with a trusted client, DRM and an encrypted source.
Your player opens up a connection, the user has a connection to the stream, you perform some magic authentication with their token and then transmite the encrypted data to them.
If you don't do this then anyone can download your video and save it out.
However with all that aside, someone can run screen capture, then save your video and do it again. This is again where the DRM comes in as one of the key features of the DRM in windows clients is that the buffer cannot be sniffed as it's on the protected media pathway.
I guess its a question of how to protect your revenue but dealing with pirates is always going to be a problem for software devs no matter what their business is.
I have a solution that i'm gonna try for myself (as I have the same worries) but I know that it includes a lot of extra time and work...
Solution: using flash compress the video into an swf file. Before compressing add some AS code to the movie for authentication. suggestions for authentication:
1 test url
2 create a dedicated flash player that has handshake code checked by the video.swf
I like #2 better, and as an extra measure, you can overlay an id code over the video, so if someone captures the video using screen recording software, you'd at least be able to track the original source of the copied video.. and exact suitable retribution...
Simply you can't prevent it.
But..you can make it difficult.
Here some ideas come in my mind
1 First of all add your identifier to the video (always someone can download it)
2 The hard way... Add Ajax call back to server to check a random generated key that it will stored in the session every N seconds. After every post back clear the buffer of the player and start the video from were i was (using javascript).
Use again JavaScript prevent the video source from downloading by "view source".
3 Handle all your videos in urls like http://www.example.com/viewvideo/1 OR ../?id=1.
Add blank image overlay with transparent background.
Serve the original video and a blank video somewhere on the page with normal extension and style attribute "display:none". (will create problems to some download helpers)
4 Everytime you serve a video CHECK if the request is from a browser (ie check UserAgent)
5 Cookie with some random value combined with the id of the video. Check it client-side and server side and then serve the video.
6 On focusout event hide the video with javascript. put a resume button in the flash and leave the frame unchange (like pause but with no original video in buffer).
7 Combine those methods
these are random generated ideas,
not tested neither i say that guaranties no video downloading.
I have attempted two way to prevent the downloading but fails.
Using javascript to dynamically generate the object for flash.
Using the token idea proposed in the question.
What annoying me most is that a simple SAVE/AS from the firefox browser could easily bypass the tricks.
The only variable way so far is to using an empty swf file to load another swf file in. Combined with the token idea, it works.
in my answer you cant stop image/video theft but you can make harder for normal users but you can't make it harder for the programmers like us( i mean thiefs that knows little web programming) there are some tricks you can try:-
1.) Using flash as youtube and many others sites like http://www.funnenjoy.com does .
2.) Div overlaping or background pic setting (but users with little sense can easily save all resources by opening inspect element or other developer option).
3.) You can disable right click and specific keys like CTRL + S and others possibles with JAVASCRIPT but main drawback is that if user disable JAVASCRIPT our all tricks fail down.
4.)Save image in none online directories(if you have full access to web server) and read that files with server side languages like PHP every time when image / video is required and change image id time to time or create script that can automatically change ID after every access.
5.)Use .htaccess in apache to prevent linking of your images by others sites. you can use this site to automatically generate .htacess http://www.htaccesstools.com/hotlink-protection/
In asp.net application, how its possible to download all png,css JavaScript and other resources parallel.
Because i am monitoring using Fiddler and found that content is downloaded one after another.
That is actually more of a browser (client) behaviour in accordance to the specification in HTTP 1.1. The guideline is to limit simultaneous downloads to two per hostname.
http://www.yuiblog.com/blog/2007/04/11/performance-research-part-4/
While you may be able to alter your browser's settings to download more per hostname, that is only your machine and not that of others' in the Internet wilderness. One way to trick clients in downloading more simulatenously is to designate your web resources into different hostnames, like images stored in http://images.yoursite.com. But you may wanna to test this and balance it out, as per the article's suggestion.
You can try AJAX for that as usually there are 5 allowed server/client http connections you could theoretically use them all at once.
However I guess you will take little advantage of this, unless you have really big (or many) css and javascript files.
Not sure if this will work on images or other files.