No permission to view FIOgroup.FabricServicesMenu

CERN FTS Services

Production

There are three FTS service instances running at CERN in production, all based on the FTS 2.1 SLC4 service. PLUS one FTS 2.2 instance used by ATLAS for T0-export production traffic.

DNS alias SLS name OS Lemon Version (20th May 2010)
fts-t2-service.cern.ch FTS-T2-SERVICE SLC4 64-bit lemon subcluster FTS 2.2.3
fts-patch4084.cern.ch FTS-PILOT-SERVICE SLC4 64-bit lemon subcluster FTS 2.2.4
fts22-t0-export.cern.ch FTS22-T0-EXPORT SLC4 64-bit lemon subcluster FTS 2.2.3

Node Deployment

This is out of date badly - 20th March 2010. gridfts-new.png

Channel deployment

This shows which FTA agent daemons are hosted upon which nodes. This is out of date badly - 20th March 2010. gridfts-channels.png

Installation and Configuration

Nodes

The FTSes are configured with Quattor, CDB NCM-YAIM and Yaim as described in FtsTier0CDBConfiguration21.

Backend database

The Oracle 10g database backend for all the FTS servers is run by IT-DM group on the LCG Oracle RAC: physics-database.support@cern.ch.

The DB connection parameters are contained in the YAIM configuration of CDB and delivered to the node via the SINDES fts_oracle_passwd cluster-level component.

CERN Site Information system

The EGEE.BDII is run as a resource BDII on all of the FTS web-servide nodes ("FTS"). The CERN-PROD site BDII should be configured to pull from these using the cluster DNS alias.

The BDII GIP running on the FTS publishes the service discovery information and the channel information (i.e. which FTS channels the service instance runs).

It is configured by the FTS module of YAIM under CDB.

 ldapsearch -x -H ldap://prod-bdii.cern.ch:2170 -b 'Mds-vo-name=CERN-PROD,o=Grid' \
'(|(|(|(|(GlueServiceType=org.glite.ChannelAgent)(GlueServiceType=org.glite.ChannelManagement))
(GlueServiceType=org.glite.Delegation))(GlueServiceType=org.glite.FileTransfer)))'

Management Procedures

All nodes are setup as standrad FIO nodes with regards to SMS, standard alarms and host certificate monitoring.

SMS states

The default SMS state is production.

  • FTS webservice: Setting a node to maintenance removes it from the load-balanced DNS alias. This should be done before any instrusive operation is performed.
  • FTA agents. Setting a node to maintenance currently has no effect on the daemons.
  • FTM monitor. Setting a node to maintenance currently has no effect on the daemons.

Monitoring

Lemon alarms

In addition to the standard FIO alarms, specific Lemon Alarms have been defined for the FTS.

Alarm name Description Comment
TOMCAT_WRONG No Tomcat processes running.  
FTA_WRONG One or more of the FTA agent daemons is down.  
FTS_STUCK The FTS web-service is not responding (although Tomcat itself is up).  
GRID_BDII_WRONG No BDII processes running.  

These alarms, along with all standard alarms on the nodes, are handled by the operator and sysadmin teams. the procedures are all stored in OPM

FTS Intervention Plans

Detailed plans for upcoming FTS interventions and records of minor interventions: FtsInterventions.

WCLG Service Review for FTS

The WLCG service review of FTS at CERN is described in FtsServiceReview20, with CERN-specific comments at FtsServiceReview20CERNPROD. They describe the impact to the overall FTS service of different failure types.

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng gridfts-channels.png r1 manage 70.8 K 2009-02-16 - 14:27 GavinMcCance FTS channel deployment
PNGpng gridfts-new.png r5 r4 r3 r2 r1 manage 100.2 K 2009-11-27 - 16:11 GavinMcCance SLC4-based FTS cluster plan
Edit | Attach | Watch | Print version | History: r36 < r35 < r34 < r33 < r32 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r36 - 2010-06-01 - SteveTraylen
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LCG All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2024 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback