Web service DRMAA interface?
Dear all, I think that this has been asked before some time back, but I don't remember seeing a definitive reply, so I will like to again ask if the WG has any plans for proposing a webservice DRMAA interface? ----------------------------------------- Melvin Koh Chee Kian Grid Research Engineer Asia Pacific Science & Technology Center Sun Microsystems Inc. Website: http://apstc.sun.com.sg/ -----------------------------------------
We've discussed it several times. The problem always comes down to how to implement the drmaa_wait() proceedure. I don't think anyone has ever proposed a solution that stuck. Anyone remeber a more concrete conclusion? Daniel Chee-Kian Koh wrote On 09/19/05 12:51,:
Dear all,
I think that this has been asked before some time back, but I don't remember seeing a definitive reply, so I will like to again ask if the WG has any plans for proposing a webservice DRMAA interface?
----------------------------------------- Melvin Koh Chee Kian Grid Research Engineer Asia Pacific Science & Technology Center Sun Microsystems Inc. Website: http://apstc.sun.com.sg/ -----------------------------------------
-- *************************************************** * Daniel Templeton ERGB01 x60220 * * Staff Engineer, Sun N1 Grid Engine * *************************************************** * "Roads? Where we're going we don't need roads." * * -Dr. Emmett Brown * * Back to the Future (1985) * ***************************************************
As stated earlier, I already spent some thoughts in these issues. One valid solution seems to be WS-Addressing (one part of WSRF). Among other things, the specification defines how you can perform an asynchronous web service call. With such a mechanism in place, the blocking drmaa_wait() call is performed by the caller through waiting for the asynchronous operation result. However, this demands clarification of what is happening if multiple interleaved operation calls appear from the same client. Considering our current time plan, I would not expect anything before the beginning of next year (Hrabri ?). Anyway it is good to know that there is interest in such a mapping. If you have a specific use case for "WS-DRMAA", please give us more details, so that we can consider your requirements in upcoming DRMAA versions. Regards, Peter. Daniel Templeton schrieb:
We've discussed it several times. The problem always comes down to how to implement the drmaa_wait() proceedure. I don't think anyone has ever proposed a solution that stuck. Anyone remeber a more concrete conclusion?
Daniel
Chee-Kian Koh wrote On 09/19/05 12:51,:
Dear all,
I think that this has been asked before some time back, but I don't remember seeing a definitive reply, so I will like to again ask if the WG has any plans for proposing a webservice DRMAA interface?
----------------------------------------- Melvin Koh Chee Kian Grid Research Engineer Asia Pacific Science & Technology Center Sun Microsystems Inc. Website: http://apstc.sun.com.sg/ -----------------------------------------
Another problem is that the current version of DRMAA does not define the interface to get the cluster info. Rayson On 9/19/05, Daniel Templeton <Dan.Templeton@sun.com> wrote:
We've discussed it several times. The problem always comes down to how to implement the drmaa_wait() proceedure. I don't think anyone has ever proposed a solution that stuck. Anyone remeber a more concrete conclusion?
Daniel
This is not as easy as it a seems. You need something like a unified monitoring terminology of (maybe) heterogeneus execution operating systems and cluster systems. The work in the CIM, JSDL and RUS standardization groups showed that this is a difficult topic. But don't worry, it is on our agenda ;-) ... Regards, Peter. Rayson Ho schrieb:
Another problem is that the current version of DRMAA does not define the interface to get the cluster info.
Rayson
On 9/19/05, Daniel Templeton <Dan.Templeton@sun.com> wrote:
We've discussed it several times. The problem always comes down to how to implement the drmaa_wait() proceedure. I don't think anyone has ever proposed a solution that stuck. Anyone remeber a more concrete conclusion?
Daniel
participants (4)
-
Chee-Kian Koh -
Daniel Templeton -
Peter Troeger -
Rayson Ho