1. Site availability Dec. 19 th 2006
Re-use materials from Piotr Nyczyk, CERN IT/GD
See presentation at last GDB

2. Service Availability Monitoring
SFT has been phased out
Service Availability Monitoring framework (SAM) :
Monitoring all grid services not only CE
It is used in the validation process of sites and services
It allows calculation of availability metrics
SAM wiki :
SAM portal :
Service and Site status are recorded (several snapshots per day)
Daily, weekly, monthly availability is calculated using integration (averaging) over the given period
Availability metrics tools in development
Official evaluation of T0 and T1 sites availability : T2 sites availability will came out soon
F. Chollet/

3. Existing tests CE, gCE job submission - UI->RB->CE->WN chain
version of CA certificates installed (on WN!)
version of software middleware (on WN!)
broker info - checking edg-brokerinfo command
UNIX shells environment consistency (BASH vs. CSH)
replica management tests - using lcg-utils, default SE defined on WN and a selected “central” SE (3-rd party replication)
accessibility of experiments software directory - environment variable, directory existence
accessibility of VO tag management tools
other tests: R-GMA client check, Apel accounting records
F. Chollet/

4. Existing tests SE, SRM LFC FTS Standalone tests
storing file from the UI - using lcg-cr command with LFC registration
getting file back to the UI - using lcg-cp command
removing file - using lcg-del command with LFC de-registration
LFC
directory listing - using lfc-ls command on /grid
creating file entry in /grid/<VO> area
FTS
checking if FTS is published correctly in the BDII
channel listing - using glite-transfer-channel-list command with ChannelManagement service
transfer test (in development):
Standalone tests
GSTAT RB
VO specific tests as well
JobWrapper tests in development… in discussion
SAM tests are not reaching all WNS
Simplified set of tests executed on WNs by a wrapper with each job grid
Core scripts installed on CEs and WNs (will become part of the realesae)
F. Chollet/

5. CE sensor Tests France Region, VO OPS
F. Chollet/

6. SE,SRM sensor Tests France Region, VO OPS
See results of put test in order to catch the problem
F. Chollet/

7. Availability metrics - algorithm∧
t ∈ CriticalTests
TestResult (N,t)
Status of node N =
Status of site S =
CE1
CE2
CEn
SRM 1
SRM 2
SRM n
site BDII
AND OR
Everything is calculated for each VO that defined critical tests in FCR
N ∈ instances(C)
Status (N)
Status of central service C = ∨
∧ = boolean AND
∨ = boolean OR
F. Chollet/

8. SAM Deployment Standalone sensors CE, gCE, SRM, FTS sensors Metrics
(GridView, XSQL)
SAM Portal
CIC ROC report
CIC on Duty Dashboard
SFT Portal
FCR
SAM Server (Oracle DB)
SFT Server (MySQL DB)
Freedom of Choice for Resources
Selection of VO Critical tests
Test definition, sites and nodes information, test results
RB monitor
GStat
SAM Submission
SAM Admin
SFT Admin
SFT Submission
Standalone sensors
CE, gCE, SRM, FTS sensors
F. Chollet/

9. Site Availability & Operation Metrics :
EGEE LCG Join effort
LCG Metrics available from SAM portal:
SAM Metrics calculation
Official evaluation of T0 and T1 sites availability :
EGEE Operations Metrics available from CIC portal (ROC Management section)
Metrics computed from SAM results and scheduled downtime
Taking into account RC (Resource Report)reports as well
F. Chollet/

10. France Region availability from EGEE CIC portal
Tool improving but site availability available as well
01/09/06 – 30/11/06
F. Chollet/

11. Availability metrics - data export
F. Chollet/

12. Availability metrics for Tiers-2
LCG Metrics
Very first attempt to export data
Non official for Tiers-2
F. Chollet/

13. To be continued…
Open issues but sites must be aware of availability metrics…
All sensors have to be reviewed and fixed:
check if tests reflect real usage
avoid dependencies on central services and third party services if possible
increase reliability of results (resistant to any other failures not related to site configuration)
increase tests verbosity (make easier to find real problem - site debugging))
Missing sensor/tests have to be written
Jobwrapper tests
Way of reviewing and fixing availability metrics
Metric calculation for aggregates of sites
F. Chollet/