ERROR:renderer_main.cc(207)] Running without renderer sandbox

[Christian LeMoussel]

Hi,

On Win 7, i started a C# console application project, I'm using
   - latest versions of the Xilium.CefGlue(xilium-xilium.cefglue-3caa551bd830)
   - Cef binary (cef_binary_3.2171.2039_windows32).

I do this CEF settings
            var cefSettings = new CefSettings
            {
                SingleProcess = false,
                MultiThreadedMessageLoop = true,
                LogSeverity = CefLogSeverity.Disable,
                LogFile = "TestOffScrennCefGlue.log"
            };

Butthis error message is displayed in console window
[0325/185611:ERROR:renderer_main.cc(207)] Running without renderer sandbox

Is this error message normal?
Why this error message is displayed in console window?

Regards,

Christian.

[Christian LeMoussel]

I find that the sandbox got introduced in CEFs 1750 branch.
I have tried setting NoSandbox = true in cefSettings and still get the same result which is kinda strange.

I find this List of Chromium Command Line Switches with --no-sandbox switch (Disables the sandbox for all process types that are normally sandboxed) but I don't know how to set this switch in xilium cefglue.

[Alex Maitland]

Safe to ignore, `CEF` will just default to running without sandbox. It really should be a warning.

[Christian LeMoussel]

Alex,

Is there a possibility to not display this message in console window?

[Alex Maitland]

Sorry, I can't answer that in the context of `CefGlue`.

[Thomas Maierhofer]

Butthis error message is displayed in console window
[0325/185611:ERROR:renderer_main.cc(207)] Running without renderer sandbox

 

Hello Christian, I'm still building CefGlue workflow elements for the Crawler-Lib Engine. I have the same issue, but my implementation will provide multiple browser runtimes, each isolated in a process.

So my child processes will just ignore this. 

[Christian LeMoussel]

Hi, Tom

How are you?

Following your message on Crawler-Lib Engine forum, I tested CefGlue to understand how it works.

Have you an idea for a date of Crawler-Lib Engine with CefGlue?

[Thomas Maierhofer]

In the next weeks. In fact it is much harder to provide headless browser workflow elements for the Crawler-Lib Engine as I thought.

Currently I'm working on multiple CEF runtimes. My first approach to use an AppDomain failed, because CEFGlue doesn't support the AppDoman Load / Unload concept. So I one of my older projects:

Child Processes (http://www.crawler-lib.net/child-processes) to start a child process for every environment you want. Like n NHunspell 32Bit and 64Bit are detected and the correct runtime will be loaded.

If you create two or more browsers in one environment, they share cookies.
If you create two browsers in different environments, they don't share cookies.

In fact the child processes for the CEFRuntime allow to perform parallel tasks if they where on different computers.

You can use different browser environments at the same time and you can configure the as you like (Cookies, Certificates Errors, Plugins, Java, JavaScript ...).

That means you can create two or more browser instances in the workflow that are truly independent because the run on a different CEF runtime. 

[Christian LeMoussel]

OK, so fun.
One question : In browser environment, can I execute a javascript and get the return value?

BR

Christian.

[Thomas Maierhofer]

The Crawler-Lib Engine workflow elements will simplify and abstract tasks, so you don't need to bother about the visitor pattern or async in CEFGlue. You will get the common workflow handlers. But I'm still doing the groundwork for this. An workflow may look like:

var runtime = await BrowserRuntime(...      

var browser = await. Browser( runtime, ...

var request = await BrowserRequest( browser, url,...

var jsResult = await BrowserJavscript( browser, ...

var image = await BrowserImage( browser, ...

var html= await BrowserHtml( browser, ...

But I haven't decided yet how to name and use the elements.

[Christian LeMoussel]

I have some ideas to test / use  Crawler-Lib Engine Browser workflow elements (in particular BrowserJavscript).
Made me know when the new version of Crawler-Lib Engine will be available with these new browser workflow elements.

[Thomas Maierhofer]

I'm highly interested in your ideas. Please post it in the Crawler_lib Engine forum under a new topic.  

Am Montag, 13. April 2015 09:56:51 UTC+2 schrieb Christian LeMoussel:

[Dmitry Azaraev]

Hi.

CefGlue itself did not write anything in console, and i'm did not know how to suppress this messages.

I'm already take some discussion in past, about this, but i'm not sure that anyone it is implemented (adds API to provide own log listener / remove existing listeners).

There is not bad idea to ask this question at ceforum.

If you ask, please sent link to topic here. Thanks.

I'm think, that it will be very useful feature, 'cause can help diagnose some specific problems.

PS: About sandbox - i'm not sure that it is really can be useful under .net, 'cause it should be initialized very early, what in fact did not happens with .net process (i.e. just after WinMain/main), so even if sanbox will work - it will be more weaky than it done in native. But if it is absolutely required, it is not too hard to create own CLR host in native, and then run own app with sandbox, but i'm also not sure that it is really work. But standard way is just doesn't use sandbox.

--
You received this message because you are subscribed to the Google Groups "CefGlue" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cefglue+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Best regards,
   Dmitry

[Dmitry Azaraev]

@Thomas Maierhofer, regardless about:

"After some research, I decide to use CefGlue instead of CefSharp for headless operations, because it supports all that low level stuff we need here.
Nevertheless this is quite hard to implement, because the Crawler-Lib Engine has to deal with different configurations and platforms (32Bit / 64Bit). In fact this is only solvable with an AppDomain. It will take a moment to implement this in a flexible way."

I'm not sure, that you need, but avoid to use CefGlue from non-default AppDomain. It will be always sucks, due by .NET AppDomain isolation (all static fields on which it rely will gone, and etc.). In same time, native CEF library should be used exactly in single-per-process manner. In same time CefGlue are can use x86 and x64 builds of CEF (at least it should work). So if you need different "execution contexts", it is better to create child-process which will host appropriate CEF. You lib itself can interoperate with it, as need. I'm already something similar, and in it good for me ('cause i'm use only CEF x86, but can operate from x64 processes).

--

Best regards,
   Dmitry

[Thomas Maierhofer]

@Dmitry: I've already switched to child processes for your mentioned issues using my own child process library:

http://www.crawler-lib.net/child-processes

For now I can start multiple CEF instances and control them as I please.