Mailing-List: contact dev-help@taverna.incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@taverna.incubator.apache.org
MIME-Version: 1.0
In-Reply-To: <11bcf975-7600-5a76-0d4c-9dd810681fde@googlemail.com>
References: <CA+bkp_=kxzmRFAdtN__0i6LSO3NLpLZh5j4MxrLcPxX7yps9Qg@mail.gmail.com>
 <11bcf975-7600-5a76-0d4c-9dd810681fde@googlemail.com>
From: Nadeesh Dilanga <nadeesh092@gmail.com>
Date: Tue, 7 Jun 2016 20:52:05 -0400
Message-ID: <CA+bkp_mpNA+6C-Dk+bwy+=9=qOBShZoKjYf4-2TcxAQU77D5Hg@mail.gmail.com>
Subject: Re: Finalize Docker Invoke JSON format
To: dev@taverna.incubator.apache.org
Content-Type: multipart/alternative; boundary=001a113de7d0ea80d70534b9b6ae
archived-at: Wed, 08 Jun 2016 00:52:11 -0000

--001a113de7d0ea80d70534b9b6ae
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi Alan,
Can you please elaborate what you mean by " It is not sufficient to just
start the container if it expects input data to be sent to stdin, or to be
in specific locations." ?

When starting a container using remote API, the only response we get is a
HTTP status code along with a message body with container id which
indicates the container started successfully. I assume you meant more on IF
the application in the container expects input data to be applied and sent
to stdin, and for such cases how we are going to cater that requirement ?


On Mon, Jun 6, 2016 at 4:33 AM, Alan Williams <alaninmcr@googlemail.com>
wrote:

> On 04-Jun-16 05:10, Nadeesh Dilanga wrote:
>
>> Hi all,
>> I am starting this thread to discuss and finalize the docker commands we
>> need to expose for client side(Taverna).
>>
>> Latest stable docker remote API is version 1.23[1]. And it has several
>> APIs
>> that can be useful.
>>
>
> Yes. Where do you intend to run the docker containers? That would be
> similar to how you can specify where to run tool services - although I
> dislike how that is done.
>
> The original JIRA [2] mentioned about the JSON format to a docker run. I
>> hope it meant about the docker config.json ?
>>
>
> That would be a lot of configuration. I am not sure that it is sensible t=
o
> have it all specified in the workflow.
>
> Because, given we use remote APIs, I would like to know what are the
>> expectations are ?
>>
>> 1. Do we assume that Images are created and published to the registry.
>>
>
> Yes, certainly for running you will know the registry and the image.
>
> 2. Do we assume that docker container is created
>>
>
> I don't think you can, as that would mean there are steps needed to be
> done before the workflow can be run.
>
> Given #1 and #2 done, then we are talking about starting the
>> container(~docker run). If that is the case, when we use remote APIs we
>> only need following, and no need of a JSON:
>>
>> Request: POST /containers/(id or name)/start
>> Response: HTTP/1.1 204 No Content
>>
>> There are other responses too:
>>
>> Status Codes:
>>
>>    - *204* =E2=80=93 no error
>>    - *304* =E2=80=93 container already started
>>    - *404* =E2=80=93 no such container
>>    - *500* =E2=80=93 server error
>>
>
> You will need to look at how the input data is read and the results
> returned. It is not sufficient to just start the container if it expects
> input data to be sent to stdin, or to be in specific locations.
>
> [1] -
>> https://docs.docker.com/engine/reference/api/docker_remote_api_v1.23/
>>
>
> Alan
>
>
>

--001a113de7d0ea80d70534b9b6ae--