April 2008 - glue-wg - lists.ogf.org

Invoking some magic: RDF, VO annotations and Grimoires
by Paul Millar 19 May '08

19 May '08

Hi all, In a round-about way, I (re)discovered a project a friend mentioned a while ago, called "Grimoires": http://twiki.grimoires.org/bin/view/Grimoires/ If you're not familiar with the project, it's basically registry that allows for service discovery, based on UDDI but allowing RDF (i.e., "semantic") annotations by multiple user communities. This should allow different end-user communities to annotate a service in almost any way supported by RDF. Re-reading the page, I was struck by what Stephen said earlier about allowing user-communities to annotate GLUE information. The interesting part comes if all GLUE entities were cast as either a UDDI-registered service or as an annotation of a service. It should be relatively easy for someone with some spare time hands (i.e., not me :) to map GLUE vocab. into RDF and allow GLUE-like publishing of information into a Grimoires server. This would allow end-users to ask some nice queries, combining their "personal" metadata with GLUE information. There's another RDF project that hook into GT4: S-MDS. IIRC, S-MDS maps the MDS information into a SPARQL end-point via some (fairly static) mapping rules. I suspect it could be persuaded to ingest GLUE (in-band via MDS or out-of-band), but 3rd-party annotations would have to be out-of-band (unless they have write-access to MDS). Whereas the Grimoires project seems to be a more integrated approach. Should we make contact with these people to see if they're interested in adopting GLUE as a case-study? It might prove useful if we provide an RDF binding for GLUE. Cheers, Paul.

2 1

Storage Sample in Glue 2.0
by Riccardo Zappi 09 May '08

09 May '08

Hi, I am spending some time to define an example of storage with the current Glue Schema draft. Unfortunately, I have some difficulties to represents the Data Store and the Storage Resource as you have defined into Glue Schema. Suppose a site with storage system organized with a SAN and cluster file system, like GPFS. Using the same SAN and cluster file system, I could define and mount different file systems upon the same infrastructure. For example, GPFS cluster can contain up to 256 mounted file systems. I can define some StorageShares using quota GPFS tools. In this case, the SS sizes are based on GPFS metadata. Therefore, I define StorageShare using file system feature. I imagine one StorageResource per each file system created on GPFS. In this case every StorageResource could have a size and quality properties. Some StorageShares could share the same StorageResource (i.e. file system) entity. When I try to describe how StorageResource is composed some doubts appears. I could lists all volumes composing the file system.. For example in the figure attached, you can see that 16 volumes compose the file system named ‘gpfs_storm’. In any case, heterogeneous disk devices could compose each volume... and then DataStore description appears incorrect. Are you sure that Storage Service administrator knows every details of underlying storage system? Moreover, if the admin knows, are you sure that he wants describe all these details? I think we have to improve DataStore-StorageResource concepts. Perhaps, we can go on to try to merge both concepts in only one entity with multi-valued quality properties. Cheers, Riccardo -- ------------------------------------------------------------------------------- Riccardo Zappi Istituto Nazionale di Fisica Nucleare - CNAF Viale Berti Pichat, 6/2, 40127 BOLOGNA - ITALY Phone: +39-051-609-2868 Fax: +39-051-609-2746 e-mail: riccardo.zappi(a)cnaf.infn.it -------------------------------------------------------------------------------

2 1

Comparison with CIM
by Paul Millar 09 May '08

09 May '08

Hi all, I don't know if anyone has been looking at this, but there's an emerging standard (through DMTF) in modelling "things" (computer-hardware and -software, network equipment, etc): CIM. There seems to be some cross-over here through OGF, so forgive me if this is already "well known". Anyway, for those who haven't heard of them, their schema makes for interesting reading. The current version (2.18) is available from: http://www.dmtf.org/standards/cim/cim_schema_v218 "Device" (from which a StorageDevice may be published) http://www.dmtf.org/standards/cim/cim_schema_v28/CIM_Device28-Final.pdf How to publish ACLs: http://www.dmtf.org/standards/cim/cim_schema_v218/CIM_Policy.pdf Enjoy! Paul.

8 24

Suggestion for splitting the StorageShare.
by Laurence Field 01 May '08

01 May '08

On further investigation of the Storage Schema an unmet use case has been found. Use Case: Consider two VOs who use two different paths to access the same StorageShare or alternatively one VO who can access the same StorageShare via two different paths. Where there could be an asymmetry with the ACLs for the SRM v2.2 space and the Path. The latest schema would require a StorageShare to be instantiated for each combination of path and SRM v2.2 space ACL, leading to n*m storage shares duplicating a great deal of information. The suggested solution is to split the StorageShare into the StorageShare and StorageEnvironment where there is a one to many relationship between the Environment and the Share. Both entities inherit the ACLs from the AccessPolicy StorageShare LocalID Path Tag StorageEnvironment LocalID ServingState AccessLatency RetentionPolicy ExpirationMode DefaultLifeTime MaximumLifeTime At the same time this allows the AggregationID to be removed.

7 16

Reminder: Glue 2.0 Schema Phone Conference
by Felix Nikolaus Ehm 28 Apr '08

28 Apr '08

Dear all, there will be a telephone conference for the Glue Schema 2.0 tomorrow morning at 11 AM CEST for the computing part: http://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneMeet ing20080429_11 and a storage part at 2 PM: http://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneMeet ing20080429 Please let me know if you can't make it. There is the possiblity of shifting the storage session later into the afternoon (3pm) if preferred. Cheers, Felix --- Felix Ehm IT-GD tel : +41 22 7674580 CERN, Switzerland -----------------------------------------

1 0

How to kill Glue! WRT OGF
by Owen Synge 28 Apr '08

28 Apr '08

Dear all, The best way to destroy Glue, is to use poorly defined objects, and untested entities relations, that are proprietary to a single or small and legacy subset of Grids, particularly with untested ideas. This will invalidate the purpose of taking your Grid specific views to the OGF. Can I please request that members of Glue trail their ideas on their own grids. I fear some members are unwilling to separate the context of their grids from the greater grid community. Examples include, binding a minority standard (SRM 2.2 and only its static spaces) into the Glue schema when numerous other examples exist such as GPFS, and NFS v4. Binding a minority authentication mechanism (VOMS) when SAML and numerous other standards are more commonly used. This may be forgivable if the people involved had tested these representations, I know from the threads on the mailing list that these ideas are new. I wish to complement Laurence on trying to stop the destruction of his work on the Glue standard, but this continued tight coupling to minority Grid services (VOMS/SRM 2.2) will lead to a legacy standard that all future grids should abandon in favour of something workable for them, if you have not even tested your ideas you should not be proposing them as a standard, that's your Grid's research work and not for OGF, which is for consolidating the Engineering objectives of interoperability. Please consider rolling out these suggestions on your own information systems AND get them WORKING before adding them to a inter grid standard. Regards Owen Synge I do not represent Dcache or DESY in my above objections, just my education as an engineer. Engineering is often about rejecting use cases so more important objectives can be achieved.

5 10

Minutes and Actions
by Laurence Field 25 Apr '08

25 Apr '08

Hi, I have added the minutes to the wiki page. http://forge.gridforum.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneM… The meeting resulted in 9 Actions, please check and let me know if anything is missing or incorrect. Thanks Laurence

1 0

Storage schema meeting during WLCG at CERN
by Felix Nikolaus Ehm 24 Apr '08

24 Apr '08

Dear all, I'd like to remind of the glue schema meeting at 2pm CEST ( in 15min). For people at CERN: We'll meet in 513-1-023 ( the computer centre, first floor - http://building.web.cern.ch/map/building?bno=513 ). Please find more details here http://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneMeet ing20080424 Cheers, Felix --- Felix Ehm IT-GD tel : +41 22 7674580 CERN, Switzerland -----------------------------------------

3 3

glue 2.0 spec comments
by Gerson Galang 23 Apr '08

23 Apr '08

Hi, I've quickly browsed through the current GLUE schema spec and want to get some clarification on the items below: 1. ComputingResource has TotalSlots and TotalLogicalCPUs. What's the difference between the two? Aren't they the same? 2. I couldn't find the Parallel attribute from ApplicationEnvironment. Have you forgotten to include this in the schema? http://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneMeeting… Thanks, Gerson

3 2

minutes of the today telecon + spec draft 40
by Sergio Andreozzi 22 Apr '08

22 Apr '08

Dear all, here you can find the minutes of the today telecon: http://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/PhoneMeeting… The update GLUE 2.0 spec (draft 40) is in the doc repository: http://forge.ogf.org/sf/docman/do/listDocuments/projects.glue-wg/docman.roo… The next telecon is next Thursday: * 24 Apr 2008 - 11 AM CEST - 1,5h - computing * 24 Apr 2008 - afternoon - F2F meeting at CERN - storage Cheers, Sergio -- Sergio Andreozzi INFN-CNAF, Tel: +39 051 609 2860 Viale Berti Pichat, 6/2 Fax: +39 051 609 2746 40126 Bologna (Italy) Web: http://www.cnaf.infn.it/~andreozzi

2 1