Hi Stephan,

I would be happy to contribute! Btw, is there an ETA on docker support with 
Mesos's Unified Containerizer ?

Best regards,
Adrian

> Le 17 mai 2016 à 17:54, Erb, Stephan <[email protected]> a écrit :
> 
> ​Excellent point.
> 
> Are you willing to contribute that patch yourself? I would happily serve as a 
> reviewer if you submit a patch and add me as a reviewer 
> (http://aurora.apache.org/documentation/latest/contributing/ 
> <http://aurora.apache.org/documentation/latest/contributing/>). 
> 
> Thanks,
> Stephan
> From: Adrian Fraisse <[email protected]>
> Sent: Tuesday, May 17, 2016 17:44
> To: [email protected]
> Subject: Re: Mesos executor dependencies using docker images
>  
> Hi Stephan,
> 
> Thank you for your reply. Knowing this, I will simply write a script to 
> automatize the update & commit of our images and it should suffice for now.
> 
> May I suggest a patch to 
> http://aurora.apache.org/documentation/latest/features/containers/ 
> <http://aurora.apache.org/documentation/latest/features/containers/> 
> mentioning Mesos dependencies though ?
> 
> Best regards,
> Adrian
> 
>> Le 17 mai 2016 à 17:35, Erb, Stephan <[email protected] 
>> <mailto:[email protected]>> a écrit :
>> 
>> Hi Adrian,
>> 
>> your analysis and solution is correct. As of today, the Aurora executor runs 
>> within the container. It is using a native mesos library in order to 
>> communicate with the agent, and therefore requires you to ship Python2.7 + 
>> all Mesos dependencies in each container.
>> 
>> Without doubt, this is not an optimal solution. We are therefore 
>> investigating alternatives: 
>> https://issues.apache.org/jira/browse/AURORA-1690 
>> <https://issues.apache.org/jira/browse/AURORA-1690>​ 
>> 
>> Best Regards,
>> Stephan
>> From: Adrian Fraisse <[email protected] 
>> <mailto:[email protected]>>
>> Sent: Tuesday, May 17, 2016 17:08
>> To: [email protected] <mailto:[email protected]>
>> Subject: Mesos executor dependencies using docker images
>>  
>> Hello fellow Aurora enthusiasts,
>> 
>> I am struggling with the docker containerizer. Jobs launched using docker 
>> containers end up failing, because Thermos cannot find Mesos's required 
>> dependencies, such as libcurl, libsasl, libsvn… See trace below, with 
>> libsasl2 : 
>> 
>> Traceback (most recent call last):
>>   File "apache/aurora/executor/bin/thermos_executor_main.py", line 45, in 
>> <module>
>>     from mesos.native import MesosExecutorDriver
>>   File 
>> "/root/.pex/install/mesos.native-0.27.2-py2.7-linux-x86_64.egg.6c53d177e078864264dbb5af282e84d7c2cdc3a8/mesos.native-0.27.2-py2.7-linux-x86_64.egg/mesos/native/__init__.py",
>>  line 17, in <module>
>>     from ._mesos import MesosExecutorDriverImpl
>> ImportError: libsasl2.so.2: cannot open shared object file: No such file or 
>> directory
>> 
>>  
>> Having stumbled upon the issue quite a lot, I found a way to fix it : having 
>> the required dependencies installed on the containers I need to run. It’s 
>> annoying and feels kinda dirty though. I dug into Mesos documentation and 
>> found this athttp://mesos.apache.org/documentation/latest/container-image/ 
>> <http://mesos.apache.org/documentation/latest/container-image/> : 
>> For a general purpose executor (e.g., thermos) of a framework (e.g., 
>> Aurora), requiring it and all its dependencies to be present in all possible 
>> container images that a user might use is not trivial.
>> 
>> Therefore, my question is : do I need to add Mesos’s dependencies to each 
>> and every docker image I want to run (and is this best practice) or am I 
>> doing something terribly wrong ?
>> 
>> Thanks for your support !
>> 
>> Adrian  
> 
> 

Reply via email to