[seek-dev] ecogrid's put service for SRB is completed

Bing Zhu bzhu at sdsc.edu
Tue Sep 14 10:32:52 PDT 2004


Jing,

I found a call in Globus to get the configuration directory, which
returns the following directory, $TOMCAT//webapps/ogsa/WEB-INF.
Please check the OpenEcogridConfigFile() in EcogridUtils.java.

Therefore, I put the code to read some configuration info such as
Metacat URL and debug flag from some configuration files
for Metacat (MetacatConfig.xml) or SRB (SrbConfig.xml) under
$TOMCAT//webapps/ogsa/WEB-INF/etc.

Please also check out the MetacatPutImpl.java for reading config
from a config file and debug message handling which I found ecogrid
(I guess Rod did it) already had a way to dealing with in EcogridUtils.java.

Bing




-----Original Message-----
From: Jing Tao [mailto:tao at nceas.ucsb.edu]
Sent: Monday, September 13, 2004 9:16 AM
To: Bing Zhu
Cc: Matt Jones
Subject: RE: [seek-dev] ecogrid's put service for SRB is completed


Hi, Bing:

I still think client user should have responsibility to assign a docid for
the put document. Ecogrid service don't need to do anything for it.

The MetacatClientTest append ".1" to a docid is because the code is client
code :) and it has to generate a docid.

The format of docid should look like bing.5.1 or bing.23.2

The first part always is username in knb(but not have to), and second part
is serial number and the last part is version number.

By the way, the url of Metacat is hard code in the implementation. Do you
think if is better to read it from a configure file?

Thanks.

Jing

On Fri, 10 Sep 2004, Bing Zhu wrote:

> Date: Fri, 10 Sep 2004 18:03:41 -0700
> From: Bing Zhu <bzhu at sdsc.edu>
> To: Jing Tao <tao at nceas.ucsb.edu>, Matt Jones <jones at nceas.ucsb.edu>
> Subject: RE: [seek-dev] ecogrid's put service for SRB is completed
>
> Hi Jing,
>
> Thanks for your comments.
>
> For the docID, I did some experiments with MatacatClient code and realized
> that the usage of '.1.' for a document name in Metacat. Actually, I
> hesitated to
> put code which appends a '.1' to a filename when uploading a file or
> inserting a
> xml document.
>
> So the question is which one of the following object ID we expect user
will
> use in calling Ecogrid 'put?
>     (1)     bing.txt.1
>     (2)     bing.txt
>
> Note that MetacatClientTest.java has code to append '.1' for each objID in
> both
> uploading a file and inserting a XML. Finally I decided to adopt this
> approach.
>
> Let me know .
>
> Bing
>
>
>
>
>
>
> -----Original Message-----
> From: Jing Tao [mailto:tao at nceas.ucsb.edu]
> Sent: Friday, September 10, 2004 3:59 PM
> To: Matt Jones
> Cc: bzhu at sdsc.edu
> Subject: Re: [seek-dev] ecogrid's put service for SRB is completed
>
>
> Hi, Bing:
>
> I took a look at the put implementation of metacat, it looks good. Thanks.
>
> But I have a issue about the docid. In uploadMetacatMetadata method, you
> generte a new id by combine docid which specified in put operation and a
> suffix - ".1". This new id will be insert into metacat. So the docid
> stored in metacat will be different to the docid which user specified in
> put operation.
>
> I guess the reseaon you want add the suffix to docid is because that if
> docid which will be inserted (not update) into metacat is not end with
> ".1", metacat will throw an exception. But I think the client user should
> have
> the responsibility to make the docid correct for metacat. How do you
> think?
>
> I am still not clear why you don't call metacatClient.update(...) method.
> If a docid user want put is bing.30.2(it is metacat data), either you
> should call update() method in metacat client library or like you do now -
> adding ".1" suffix to "bing.30.2" and call insert(). As I mentioned before
> the id stored in metacat will be bing.30.2.1 and it is different to
> bing.30.2
>
> If without the change of metacat implementation, I think we have to use
> update in metacat client library to handle the docid which is not revision
> one.
>
> By the way, above discussion only focus on local indentifier rather than
> LSID indentifier.
>
> This is my two cents :)
>
> Jing
>
>
> On Fri, 10 Sep 2004, Matt Jones wrote:
>
> > Date: Fri, 10 Sep 2004 07:15:44 -0800
> > From: Matt Jones <jones at nceas.ucsb.edu>
> > To: bzhu at sdsc.edu
> > Cc: Jing Tao <tao at nceas.ucsb.edu>
> > Subject: Re: [seek-dev] ecogrid's put service for SRB is completed
> >
> > Hi Bing,
> >
> > No, I didn't mean that.  The metacat update command is superfluous given
> > our decision to not allow revisions in the identifiers.  Therefore, I
> > think there should be three operations:
> >  >    login(username, passwd) returns sessionID
> >  >    put(...)  as designed
> >  >    logout(sessionID)
> >
> > This means that, depending on the ID, in metacat some put operations
> > would call Metacat.insert(...) and others would call Metacat.upload(...)
> > depending on whether its metadata or data that is being put.  There
> > might even be a reason to call Metacat.update(...) from the put routine,
> > but I don't think so.
> >
> > Also, these same three operations should work identically for SRB and
> > Metacat (ie, no signature differences there, right?).
> >
> > Can you ask Jing to take a look at your implementation when you're done
> > testing?  Have you developed some junit test for testing?  You could
> > probably modify the MetacatClientTest.java class pretty easily to do
> > Ecogrid calls instead in testing.
> >
> > Matt
> >
> > Bing Zhu wrote:
> > > Matt,
> > >
> > > I finished the coding 'put' for Metacat and checked in the code.
> > > Currently I am testing the service.
> > >
> > > When I reviewed you mail below this morning, I wonder if you mean
> > > there should be four operations in put service.
> > >    login(username, passwd) returns sessionID
> > >    put(...)  as designed
> > >    update() similar to put()
> > >    logout(sessionID)
> > >
> > > Note that put actually call 'upload' or 'insert' depending the data
> type.
> > >
> > > Bing
> > >
> > >
> > >
> > >
> > >
> > > -----Original Message-----
> > > From: Matt Jones [mailto:jones at nceas.ucsb.edu]
> > > Sent: Monday, August 23, 2004 1:25 PM
> > > To: bzhu at sdsc.edu
> > > Cc: Jing Tao; Saurabh Garg; Bertram Ludaescher
> > > Subject: Re: [seek-dev] ecogrid's put service for SRB is completed
> > >
> > >
> > > Hi Bing,
> > >
> > > Welcome back.
> > >
> > > The 'get' implementation for metacat's ecogrid interface uses our
> > > metacat client library. So, I think 'put' should be implemented in a
> > > parallel way.   To do so, you need metacat-client.jar, not
metacat.jar.
> > >   The easiest way for you to get everything that you need is to check
> > > out a copy of the 'metacat' and the 'utilities' module out of cvs.
Your
> > > current SEEK cvs information should work fine with a command like
this:
> > >
> > > cvs co -P metacat
> > > cvs co -P utilities
> > >
> > > Then, to get the metacat-client.jar, type 'ant clientjar' from within
> > > the metacat directory, which will create the jar for you in the build
> > > directory.
> > >
> > > The metacat API has three interfaces that are relevant to your 'put'
> > > implementation:
> > >    insert,  update, and upload
> > > You will also need to utilize the 'login' interface to establish
> > > identity before being able to write to a service.
> > >
> > > These are all defined in the interface file:
> > >    metacat/src/edu/ucsb/nceas/metacat/client/Metacat.java
> > >
> > > You and Bertram and I need to discuss your activities for the next few
> > > weeks -- Bertram and I came up with some possibly higher priority work
> > > on the ecogrid registry, but I'd like to talk to you and Bertram about
> > > it for a bit to decide how to prioritize stuff.   Can we do that this
> > > week?  Thanks,
> > >
> > > Matt
> > >
> > > Saurabh Garg wrote:
> > >
> > >>Hi Bing,
> > >>
> > >>I am not sure about the particulars of put operation in Ecogrid. Hence
I
> > >>dont know about the format of the parameter 'object'.
> > >>
> > >>Session id is the id you get when you log into metacat. When you log
> into
> > >>metacat, the xml that is sent back contains the session id - the id
that
> > >
> > > is
> > >
> > >>used to identify you (the user) during the life of the session. Once
you
> > >
> > > log
> > >
> > >>out, the session id is deleted from the server. So to get that session
> id,
> > >>you will have to do a login first.
> > >>
> > >>The most recent jar file is attached. (I hope you were refering to
this
> > >
> > > file
> > >
> > >>from the build directory and not an install jar for metacat)
> > >>
> > >>Hope that helps,
> > >>Sid
> > >>
> > >>----- Original Message -----
> > >>From: "Bing Zhu" <bzhu at sdsc.edu>
> > >>To: "Matt Jones" <jones at nceas.ucsb.edu>; <sgarg at nceas.ucsb.edu>
> > >>Cc: "Bertram Ludaescher" <ludaesch at sdsc.edu>; "Arcot Rajasekar"
> > >><sekar at sdsc.edu>; "Jing Tao" <tao at nceas.ucsb.edu>; "Seek-Dev"
> > >><seek-dev at ecoinformatics.org>
> > >>Sent: Monday, August 23, 2004 12:03 PM
> > >>Subject: RE: [seek-dev] ecogrid's put service for SRB is completed
> > >>
> > >>
> > >>
> > >>
> > >>>Matt and Sid,
> > >>>
> > >>>I am back to my office and would like to start implementation of
> Ecogrid
> > >>>'put' for
> > >>>Metacat. Would you send me the followings.
> > >>>
> > >>>1. APIs used for uploading a Metacat document or the metadata for a
> > >>
> > >>Metacat
> > >>
> > >>
> > >>>document.
> > >>>
> > >>>       Basically the 'put' operation in Ecogrid's put service has
> > >>
> > >>following
> > >>
> > >>
> > >>>form.
> > >>>              put(object, objectId, type, sessionId)
> > >>>
> > >>>       The type can be either 'data' or 'metadata'.
> > >>>       In Metacat case, I think, the 'object' will be a byte array
> which
> > >>
> > >>is
> > >>
> > >>
> > >>>actually
> > >>>       the XML string for a XML document or metadata.
> > >>>
> > >>>2. The usage of  "session ID" in Metacat's uploading function.
> > >>>3. Latest Metacat jar file.
> > >>>
> > >>>Thanks.
> > >>>
> > >>>Sincerely,
> > >>>Bing
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>-----Original Message-----
> > >>>From: seek-dev-admin at ecoinformatics.org
> > >>>[mailto:seek-dev-admin at ecoinformatics.org]On Behalf Of Jing Tao
> > >>>Sent: Friday, August 13, 2004 4:12 PM
> > >>>To: Bing Zhu
> > >>>Cc: Matt Jones; Bertram Ludaescher; Arcot Rajasekar; Seek-Dev;
> > >>>sgarg at nceas.ucsb.edu
> > >>>Subject: Re: [seek-dev] ecogrid's put service for SRB is completed
> > >>>
> > >>>
> > >>>Hi, bing:
> > >>>
> > >>>Great! It is glad to know the implementation is done for srb.
> > >>>
> > >>>Unfortunately I will be off from Aug. 23 to Aug. 27. If you have some
> > >>>thing about metacat, you may talk with matt or sid (who is in charge
> > >>>metacat now and his email is sgarg at nceas.ucsb.edu)
> > >>>
> > >>>Have a good vaction!
> > >>>
> > >>>Jing
> > >>>
> > >>>On Fri, 13 Aug 2004, Bing Zhu wrote:
> > >>>
> > >>>
> > >>>
> > >>>>Date: Fri, 13 Aug 2004 15:31:27 -0700
> > >>>>From: Bing Zhu <bzhu at sdsc.edu>
> > >>>>To: Matt Jones <jones at nceas.ucsb.edu>
> > >>>>Cc: Bertram Ludaescher <ludaesch at sdsc.edu>, Arcot Rajasekar
> > >>>
> > >>><sekar at sdsc.edu>,
> > >>>
> > >>>
> > >>>>    Jing Tao <tao at nceas.ucsb.edu>, Seek-Dev
> > >>
> > >><seek-dev at ecoinformatics.org>
> > >>
> > >>>>Subject: [seek-dev] ecogrid's put service for SRB is completed
> > >>>>
> > >>>>Matt,
> > >>>>
> > >>>>The 'put' service for SRB has been implemented and tested. I have
> > >>
> > >>checked
> > >>
> > >>
> > >>>in
> > >>>
> > >>>
> > >>>>all code, stubs, SRB implementation code and a ecogrid put client.
> > >>>>
> > >>>>I will take next week off for a vacation break and will contact you
> and
> > >>>
> > >>>Jing
> > >>>
> > >>>
> > >>>>for the Metacat implementation after I come back (Aug 23).
> > >>>>
> > >>>>Sincerely,
> > >>>>Bing
> > >>>>
> > >>>>
> > >>>>BWT,
> > >>>>The SRB implementation files:
> > >>>>src\org\ecoinformatics\ecogrid\srb\put\SrbPutImpl.java
> > >>>>
> > >>>
> > >>>src\org\ecoinformatics\ecogrid\srb\SRBuri.java
> > >>>
> > >>>
> > >>>>Ecogrid put client:
> > >>>>src\org\ecoinformatics\ecogrid\client\EcogridPutClient.java
> > >>>>Ecogrid Put stubs from gwsdl:
src\org\ecoinformatics\ecogrid\stub\...
> > >>>>
> > >>>>Currently the put service for SRB system is running in the test
> machine:
> > >>>>
> > >>>>
> > >>>
> > >
>
http://sasa.sdsc.edu:8080/ogsa/services/org/ecoinformatics/ecogrid/EcoGridLe
> > >
> > >>>>velOnePutService
> > >>>>  (for web browser: http://sasa.sdsc.edu:8080/ogsa/services)
> > >>>>
> > >>>>_______________________________________________
> > >>>>seek-dev mailing list
> > >>>>seek-dev at ecoinformatics.org
> > >>>>http://www.ecoinformatics.org/mailman/listinfo/seek-dev
> > >>>>
> > >>>
> > >>>--
> > >>>Jing Tao
> > >>>National Center for Ecological
> > >>>Analysis and Synthesis (NCEAS)
> > >>>735 State St. Suite 204
> > >>>Santa Barbara, CA 93101
> > >>>
> > >>>_______________________________________________
> > >>>seek-dev mailing list
> > >>>seek-dev at ecoinformatics.org
> > >>>http://www.ecoinformatics.org/mailman/listinfo/seek-dev
> > >>
> > >
> > > --
> > > -------------------------------------------------------------------
> > > Matt Jones                                     jones at nceas.ucsb.edu
> > > http://www.nceas.ucsb.edu/    Fax: 425-920-2439    Ph: 907-789-0496
> > > National Center for Ecological Analysis and Synthesis (NCEAS)
> > > University of California Santa Barbara
> > > Interested in ecological informatics? http://www.ecoinformatics.org
> > > -------------------------------------------------------------------
> >
> >
>
> --
> Jing Tao
> National Center for Ecological
> Analysis and Synthesis (NCEAS)
> 735 State St. Suite 204
> Santa Barbara, CA 93101
>
>

--
Jing Tao
National Center for Ecological
Analysis and Synthesis (NCEAS)
735 State St. Suite 204
Santa Barbara, CA 93101




More information about the Seek-dev mailing list