Adding service endpoints

Adding service endpoints information into Configuration Database

For monitoring purposes, each service endpoints registered into the Configuration Database, and having the flags production and monitored should include the endpoint URL information in order to be contacted by the corresponding service-specific nagios probe.

The information needed for service type are:

  • SRM: the value of the attribute GlueServiceEndpoint published in the Configuration Database or BDII (e.g. httpg://se.egi.eu:8444/srm/managerv2)
  • Cloud:
    • org.openstack.nova: The endpoint URL must contain the Keystone v3 URL: https://hostname:port/url/v3
    • org.openstack.swift:The endpoint URL must contain the Keystone v3 URL: https://hostname:port/url/v3
    • eu.egi.cloud.accounting: for the host sending the records to the accounting repository
  • Other service types: the value of the attribute GlueServiceEndpoint published in the BDII

It is also possible to register additional endpoints for every services, they will also be monitored if the “Monitored” flag is set.

For having more information about managing the Service endpoints in the Configuration Database, please consult the service endpoints documentation.

Retrieving the information

For retrieving the queue URL from the BDII, you can use the command lcg-infosites, to be executed from a UI. Be sure to query a production Top BDII: you can either use the one provided by your Operations Centre or choose one from the Configuration Database

For example:

$ export LCG_GFAL_INFOSYS=egee-bdii.cnaf.infn.it:2170

$ lcg-infosites --vo ops ce | grep nikhef
  5680      15          0            0       0 dissel.nikhef.nl:2119/jobmanager-pbs-infra
  5680      17          1            1       0 gazon.nikhef.nl:8443/cream-pbs-infra
  5680      15          0            0       0 juk.nikhef.nl:8443/cream-pbs-infra
  5680      15          0            0       0 klomp.nikhef.nl:8443/cream-pbs-infra
  5680      16          0            0       0 stremsel.nikhef.nl:8443/cream-pbs-infra

In order to find the GlueServiceEndpoint URL of your SRM service, you can launch a LDAP query to your Site BDII (or directly to the SRM service):

$ ldapsearch -x -LLL -H ldap://sbdii01.ncg.ingrid.pt:2170 \
  -b "mds-vo-name=NCG-INGRID-PT,o=grid" \
  '(&(objectClass=GlueService)(GlueServiceType=SRM))' \
  GlueServiceEndpoint

dn: GlueServiceUniqueID=httpg://srm01.ncg.ingrid.pt:8444/srm/managerv2,Mds-Vo-name=NCG-INGRID-PT,o=grid
GlueServiceEndpoint: httpg://srm01.ncg.ingrid.pt:8444/srm/managerv2

In a similar way, by just changing the value of GlueServiceType, you can retrieve the endpoint URLs of other services.

An alternative way for retrieving the GlueServiceEndpoint URL is using the GLUE2 information browser provided by VAPOR: select your NGI, then your site and hence the Storage service; click on the endpoint details button for finding the URL associated to the SRM interface.

Filling the information in

URLs information are completely missing

Editing the services information

  • Site overview

This is the home page regarding your site. You need to fill in the URL information.

Click on a service for displaying its page (e.g. the CREAM-CE).

Site overview

  • Editing a service

Click on the EDIT button in the top right corner

Service overview

  • Adding a Service URL

fill in the Service URL field with the queue URL

Service Edition

  • Reviewing the site

Now the CREAM-CE service endpoint contains the required queue information.

Proceed in a similar way for the other services.

Completed service overview

Additional endpoints information

In case you need to register an additional endpoint for a service, go on the service summary page and add the proper information. In the example below it is shown the case of a computing element.

  • Service summary page

This is the service summary page.

You need to click on the Add endpoint button for registering additional endpoint URLs.

Service summary page

  • Adding an endpoint

Fill in the proper information and don’t forget to select the “Monitored” flag for making Nagios to detect the new endpoint.

Adding and endpoint

  • Reviewing the endpoint description

The summary page of the endpoint just added should look like this one.

Service endpoint page

  • Reviewing the service description

And this is the summary page of the service reporting the information about all its endpoints registered: the first one in the Grid Information section and the additional ones in the Service Endpoints section.

Service summary page

Examples

webdav

In order to properly monitor your webdav endpoint:

  • you should register a new service endpoint with the webdav service type, separated from the SRM one;
  • the endpoint URL information used for monitoring purposes should be set in the extension properties section. Create the following:
    • Name: ARGO_WEBDAV_OPS_URL
    • Value: webdav URL containing also the VO ops folder, for example: https://darkstorm.cnaf.infn.it:8443/webdav/ops or https://hepgrid11.ph.liv.ac.uk/dpm/ph.liv.ac.uk/home/ops/
      • it corresponds to the value of GLUE2 attribute GLUE2EndpointURL (containing the used port and without the VO folder);
  • verify that the webdav URL (for example: https://darkstorm.cnaf.infn.it:8443/webdav) is properly accessible.

EOS and XrootD service endpoints

The EOS service endpoints expose an XrootD interface, so in order to properly monitor them, even in case you provide a plain XrootD endpoint, please do the following:

  • you should register a new service endpoint with the XrootD service type;
  • the endpoint URL information used for monitoring purposes should be set in the extension properties section. Create the following:
    • Name: ARGO_XROOTD_OPS_URL
    • Value: XRootD base SURL to test (the path where ops VO has write access), for example: root://eosatlas.cern.ch//eos/atlas/opstest/egi/, root://recas-se-01.cs.infn.it:1094/dpm/cs.infn.it/home/ops/, root://dcache-atlas-xrootd-ops.desy.de:2811/pnfs/desy.de/ops or similar). Pay attention to the port configured for the interface.

GridFTP

In order to properly monitor your gridftp endpoint for ops VO

  • register a new service endpoint, associating the storage element hostname to the service type globus-GRIDFTP, with the “production” flag disabled;
  • in the “Extension Properties” section of the service endpoint page, fill in the following fields:
    • Name: SE_PATH
    • Value: /dpm/ui.savba.sk/home/ops (this is an example, set the proper path)
  • check if the tests are OK (it might take some hours for detecting the new service endpoint) and then switch the production flag to “yes”

SURL value for SRM

The SURL value needed by the SRM monitoring probes is the following structure:

srm://<hostname>:<port>/srm/managerv2?SFN=<GlueSAPath or GlueVOInfoPath>

For example:

srm://ccsrm.in2p3.fr:8443/srm/managerv2?SFN=/pnfs/in2p3.fr/data/dteam/

  • As explained in previous sections, you can retrieve the port number from the GlueServiceEndpoint URL information.
  • If your SE provides GlueSAPath information, use that. To retrieve it:
$ ldapsearch -x -LLL -H <ldap://sbdii01.ncg.ingrid.pt:2170> \
  -b "mds-vo-name=NCG-INGRID-PT,o=grid" \
  '(&(objectClass=GlueSA)(GlueSAAccessControlBaseRule=VO:ops))' \
  GlueSAPath GlueChunkKey

dn: GlueSALocalID=opsdisk:replica:online,GlueSEUniqueID=srm01.ncg.ingrid.pt,Mds-Vo-name=NCG-INGRID-PT,o=grid
GlueChunkKey: GlueSEUniqueID=srm01.ncg.ingrid.pt
GlueSAPath: /gstore/t2others/ops
  • if your SE doesn’t provide GlueSAPath information, use instead the GlueVOInfoPath one:
$ ldapsearch -x -LLL -H <ldap://ntugrid5.phys.ntu.edu.tw:2170> \
  -b "Mds-Vo-name=TW-NTU-HEP,o=grid" \
  (&(objectClass=GlueVOInfo)(GlueVOInfoAccessControlBaseRule=VO:ops)) \
  GlueVOInfoLocalID GlueChunkKey

dn: GlueVOInfoLocalID=ops:SRR,GlueSALocalID=SRR:SR:replica:*****,GlueSEUniqueID=ntugrid6.phys.ntu.edu.tw,Mds-Vo-name=TW-NTU-HEP,o=grid
GlueVOInfoPath: /dpm/phys.ntu.edu.tw/home/ops
GlueChunkKey: GlueSALocalID=SRR:SR:replica:*****
GlueChunkKey: GlueSEUniqueID=ntugrid6.phys.ntu.edu.tw
GlueVOInfoLocalID: ops:SRR

dn: GlueVOInfoLocalID=ops:data01,GlueSALocalID=data01:replica:online,GlueSEUniqueID=ntugrid6.phys.ntu.edu.tw,Mds-Vo-name=TW-NTU-HEP,o=grid
GlueVOInfoPath: /dpm/phys.ntu.edu.tw/home/ops
GlueChunkKey: GlueSALocalID=data01:replica:online
GlueChunkKey: GlueSEUniqueID=ntugrid6.phys.ntu.edu.tw
GlueVOInfoLocalID: ops:data01
  • Pay attention to use the storage path for the ops VO
  • On GOCDB, in the “Extension Properties” section of the SRM service endpoint page, fill in the following fields:
    • Name: SURL
    • Value: the actual SURL value, for example: srm://srm01.ncg.ingrid.pt:8444/srm/managerv2?SFN=/gstore/t2others/ops