how to publish services information
Dear all, In reference to the discussion during our last phone meeting I found that GLUE seems not to inforce any standar name. But I also found a rather document (from 2009. You can find it attached to this email) that describe the name usage in WLCG. In particular, concerning the storage: * ImplementationName MUST be published. The format is defined by the implementations, but the names MUST NOT clash. In particular, for WLCG storage services the names to be used are: CASTOR, DPM, StoRM, dCache, BeStMan, xrootd. * ImplementationVersion MUST be published. The format is defined by the implementations. For OS names there are others directives here (I don not think we need it but while I was looking around for info I found it): https://wiki.egi.eu/wiki/Operations/HOWTO05 For site names: https://wiki.egi.eu/wiki/Operations/HOWTO05 Furthermore my understanding is that, at least in EMI, the decision is just to set the EMI release version. Based on that it should be then possible to identify the version of the specific middleware. Cheers, Andrea
Hi, I think it may be interesting to know the implementation but I am struggling to see why it is a MUST and also how we would aim for the MUST NOT Clash. If it is used I agree a project should themselves ensure that no clashes occur but only the infrastructure/project using the standard can define their resource types and ensure this. can someone help clear this up for me - I may have missed some of the background discussion? cheers johnk On 06/15/2012 10:03 AM, Andrea Cristofori wrote:
Dear all,
In reference to the discussion during our last phone meeting I found that GLUE seems not to inforce any standar name. But I also found a rather document (from 2009. You can find it attached to this email) that describe the name usage in WLCG. In particular, concerning the storage:
* ImplementationName MUST be published. The format is defined by the implementations, but the names MUST NOT clash. In particular, for WLCG storage services the names to be used are: CASTOR, DPM, StoRM, dCache, BeStMan, xrootd.
* ImplementationVersion MUST be published. The format is defined by the implementations.
For OS names there are others directives here (I don not think we need it but while I was looking around for info I found it):
https://wiki.egi.eu/wiki/Operations/HOWTO05
For site names:
https://wiki.egi.eu/wiki/Operations/HOWTO05
Furthermore my understanding is that, at least in EMI, the decision is just to set the EMI release version. Based on that it should be then possible to identify the version of the specific middleware.
Cheers, Andrea
-- ur-wg mailing list ur-wg@ogf.org https://www.ogf.org/mailman/listinfo/ur-wg
-- +------------------------------------------------------------+ |Dr. John Alan Kennedy Rechenzentrum Garching (RZG) | |Mail: jkennedy@rzg.mpg.de Boltzmannstrasse 2 | |Phone: +49 89 3299 2694 85748 Garching | |Fax: +49 89 3299 1301 | +------------------------------------------------------------+
John, the background discussion during the last meeting was whether there was a place to register these values to lookup what you should use and avoid clashes. It was thought GLUE might provide a place but it seems not. I don't think we want to build the list into the standard but OGF could provide a place on its wiki where a list of known ones is maintained. Regards, John From: ur-wg-bounces@ogf.org [mailto:ur-wg-bounces@ogf.org] On Behalf Of john alan kennedy Sent: 15 June 2012 10:39 To: Andrea Cristofori Cc: ur-wg@ogf.org Subject: Re: [UR-WG] how to publish services information Hi, I think it may be interesting to know the implementation but I am struggling to see why it is a MUST and also how we would aim for the MUST NOT Clash. If it is used I agree a project should themselves ensure that no clashes occur but only the infrastructure/project using the standard can define their resource types and ensure this. can someone help clear this up for me - I may have missed some of the background discussion? cheers johnk On 06/15/2012 10:03 AM, Andrea Cristofori wrote: Dear all, In reference to the discussion during our last phone meeting I found that GLUE seems not to inforce any standar name. But I also found a rather document (from 2009. You can find it attached to this email) that describe the name usage in WLCG. In particular, concerning the storage: * ImplementationName MUST be published. The format is defined by the implementations, but the names MUST NOT clash. In particular, for WLCG storage services the names to be used are: CASTOR, DPM, StoRM, dCache, BeStMan, xrootd. * ImplementationVersion MUST be published. The format is defined by the implementations. For OS names there are others directives here (I don not think we need it but while I was looking around for info I found it): https://wiki.egi.eu/wiki/Operations/HOWTO05 For site names: https://wiki.egi.eu/wiki/Operations/HOWTO05 Furthermore my understanding is that, at least in EMI, the decision is just to set the EMI release version. Based on that it should be then possible to identify the version of the specific middleware. Cheers, Andrea -- ur-wg mailing list ur-wg@ogf.orgmailto:ur-wg@ogf.org https://www.ogf.org/mailman/listinfo/ur-wg -- +------------------------------------------------------------+ |Dr. John Alan Kennedy Rechenzentrum Garching (RZG) | |Mail: jkennedy@rzg.mpg.demailto:jkennedy@rzg.mpg.de Boltzmannstrasse 2 | |Phone: +49 89 3299 2694 85748 Garching | |Fax: +49 89 3299 1301 | +------------------------------------------------------------+ -- Scanned by iCritical.
dear all, On 2012-06-15 12:21, john.gordon@stfc.ac.uk wrote:
John, the background discussion during the last meeting was whether there was a place to register these values to lookup what you should use and avoid clashes. It was thought GLUE might provide a place but it seems not. I don’t think we want to build the list into the standard but OGF could provide a place on its wiki where a list of known ones is maintained.
sorry for interrupting but let me point your attention to the rather recent GLUE2 work of service type/name enumeration cleanup. This will be on the agenda at the GLUE2 session in Delft (Sunday afternoon). excerpt from GLUE2 wiki page: https://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/HomePage A review of Open Enumerations is ongoing. The current results for ServiceType have been updated here: https://docs.google.com/spreadsheet/ccc?key=0An1CLx-Ye4OFdDkxWmxjNkZkS0xrc0t... the work in progress is being shaped here: https://docs.google.com/spreadsheet/ccc?key=0An1CLx-Ye4OFdDRkV180clZQampoRGw... - access to the doc can be requested on the wg mailing list. A google account is needed. regards, Balazs Konya as glue2 co-chair -- Balázs Kónya Technical Director European Middleware Initiative www.eu-emi.eu NorduGrid Collaboration www.nordugrid.org Lund University balazs.konya@hep.lu.se Institute of Physics, EHEP phone: +46 46 222 8049 BOX 118, S - 221 00 LUND, Sweden fax: +46 46 222 4015
Dear Balazs, thank you for the input. We will have a look at it for sure. Your session at OGF is just before so I will try to be there. BTW, maybe there is a misconfiguration because I can see the doc on the google page. Regards, Andrea On 06/15/2012 01:24 PM, Balazs Konya wrote:
dear all,
On 2012-06-15 12:21, john.gordon@stfc.ac.uk wrote:
John, the background discussion during the last meeting was whether there was a place to register these values to lookup what you should use and avoid clashes. It was thought GLUE might provide a place but it seems not. I don’t think we want to build the list into the standard but OGF could provide a place on its wiki where a list of known ones is maintained.
sorry for interrupting but let me point your attention to the rather recent GLUE2 work of service type/name enumeration cleanup. This will be on the agenda at the GLUE2 session in Delft (Sunday afternoon).
excerpt from GLUE2 wiki page: https://forge.ogf.org/sf/wiki/do/viewPage/projects.glue-wg/wiki/HomePage
A review of Open Enumerations is ongoing. The current results for ServiceType have been updated here:
https://docs.google.com/spreadsheet/ccc?key=0An1CLx-Ye4OFdDkxWmxjNkZkS0xrc0t...
the work in progress is being shaped here: https://docs.google.com/spreadsheet/ccc?key=0An1CLx-Ye4OFdDRkV180clZQampoRGw... - access to the doc can be requested on the wg mailing list. A google account is needed.
regards, Balazs Konya
as glue2 co-chair
-- Andrea Cristofori INFN-CNAF Viale Berti Pichat 6/2 40127 Bologna Italy Tel. : +39-051-6092920 Skype: andrea-cnaf
participants (4)
-
Andrea Cristofori -
Balazs Konya -
john alan kennedy -
john.gordon＠stfc.ac.uk