Thank you Alan!
I thought the Taverna Server Virtual Machine was a
Taverna 2
server...
(http://www.mygrid.org.uk/taverna/internal/taverna-server/2.2a1/taverna-server-2.2a1-demonstrator-vm-ovf-2010-07-30.zip).
So if I place this machine on the internet it would not be possible to
use it
as (personal) runner machine to run my experiments from Myexperiment
directly on
it.
Is that correct? I have to setup a Taverna 1
server to make
use of this runner-feature in MyExperiment?
In MyExperiment the API coverage it’s noted that
GET runner
is implemented (http://wiki.myexperiment.org/index.php/Developer:APICoverage).
Where
can I find any information on this? What does it return?
What we would like to do is:
- User is using a MyExperiment-gadget in its own
portal (possible).
- User has to authenticate to use the
MyExperiment-gadget (possible)..
- The MyExperiment gadget lists the workflows of
the user (possible).
- The MyExperiment gadget also lists the available
runners
for that user (is this already possible at the moment when pre
configured in MyExperiment).
- The user has the option to run one of these workflows on a available
runner
(is this possible at the moment making use of the current API through
the gadget).
- The user should have the opportunity to see the status of his running
experiment in this gadget (not possible as I understand form your
answer?).
To make Runners available for running workflows we
would
like to setup a Taverna server, that’s why we need to know the version
we can
use. And how the RUNNER API works.
Doe you recommend contacting somebody directly to
see if
this scenario is possible?
Kind regards and thank you for the quick response,
Wladimir
On 10-11-2010 13:13, Alan R Williams wrote:
On
10/11/2010 11:57, Wladimir Mufty wrote:
Hello MyExperiment users and developers,
Hello Wladimir,
[snip]
Unfortunately I could not attach a pdf with
the scenario to this
mailinglist, or should this be possible?
There are still some questions we have:
- Is the runner that can be configured in MyExperiment pointing to an
instance of a Taverna server?
Yes, but currently a Taverna 1 server. Some work will need to be done
to have a Taverna 2 runner.
Can I use the Taverna Server Virtual
machine for this when I place it on my (public) network?
You should be able to.
- When “posting” an experiment to a
runner-machine through the
MyExperiment API, does MyExperiment does that directly, or should I
first GET the workflow and POST it manually to the runner.
The only difference I can see would be for authentication of the user.
- When do I need to use the concept of a job?
I don't understand the question.
- Can I check the status of a experiment
(posted to a MyExperiment
runner) through the MyExperiment API
As far as I know, nothing about the runners is exposed in the
myExperiment APIs
or directly from that machine it’s
running on?
Yes. There are REST and WSDL interfaces to the Taverna 2 Server. The
Ruby demonstrator uses the REST API to talk to the server.
Is there some (draft) documentation available
on the following API
subjects; we would like to make use of them… even if there not official
yet.
- GET group, POST group, PUT group, DELETE group
The myExperiment REST read API is documented at
http://wiki.myexperiment.org/index.php/Developer:API
Someone from myExperiment should be able to tell you where the write
API is.
- Runner index , Runner count, GET runner,
POST runner, PUT runner,
DELETE runner
Those are not available.
- Job index , Job count, GET Job, POST Job,
PUT Job, DELETE Job
I can ask Donal to send you the URL to the REST API for Taverna 2
Server. We are having a meeting today about changing how it handles
lists.
Hope you could give me some input on these
questions,
Kind regards,
Wladimir Mufty
Alan
_______________________________________________
Myexperiment-discuss mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/myexperiment-discuss