This table contains examples of the different availabilities in SAM2 and SAM3, the explanation for it, and any comments from the experiments and operations representatives.

The possible status for each entry are:

  • Under investigation: it is not clear what causes this difference
  • Understood: the issue has been identified, and it has to be agreed that the SAM3 behaviour is the expected one
  • Need fixing: The issue has to be fixed in SAM3
  • Solved: the issue has been undersood, and the SAM3 behaviour is the expected

Instructions

If you are an experiment or operations representative, please read the 'understood' rows of the table carefully, and put a Green led if you agree with the explanation and this is the behaviour that you expect in SAM3, or a comment if you don't agree.

T0/T1 differences

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL   No issues here here             Understood  
ATLAS ATLAS_CRITICAL Taiwan, RRC-KI Wrong combination of SRM instances click here click here SAM2 seems to do an 'OR' of the SRM instances. SAM3 is doing the AND (according to the algorithm definition)   Green led       Understood &SRM
ATLAS ATLAS_CRITICAL FZK Missing maintenance here here The downtime took 3 hours. SAM2 presents the whole entry as brown. SAM3 presents the availability of the entry (0.88), and a small white fraction that represents the downtime (0.12)           Understood Display
ATLAS ATLAS_CRITICAL RAL Unknown service --> Available Site click here click here If all instances of a services are unknown, SAM2 does not take the instance into account for the site availability. SAM 3 passes the unknown to the site availability           Understood Unknown service ignored
LHCb LHCb_CRITICAL All sites Different metrics in profiles   here The metrics included in the algorithm for the CREAM-CE were different. Removing two obsolete metrics from SAM3, and recalculating the availability           Solved  
LHCb LHCB_CRITICAL NIKHEF.nl, RAL.uk Sites do not have SRM, availability in SAM3 unknown here here The sites do not have an srm defined in the vofeed. SAM2 doesn't complain, SAM3 marks the availability as unknown. Update: the logic of 'and' has been modified to 'and if data''. There should not be differences here           Understood Logic of 'and'
CMS CMS_CRITICAL_FULL T0_CH_CERN, T1_CH_CERN Sites do not have any services   here The same service needs to appear in multiple sites.           Solved  
CMS CMS_CRITICAL_FULL All sites Different metrics in profile   here The metrics included in the algorithm for the CREAM-CE were different. Removing cvmfs , and recalculating the availability           Solved  
CMS CMS_CRITICAL_FULL T1_UK_RAL_Disk, T1_RU_JINR_Disk Sites appear in SAM3 and not in SAM2 here here The sites appear in the cms vofeed, so they are passes to SAM3. SAM2 filters them out because the sites do not appear in GOCDB     Green led     Understood New service in vofeed
CMS CMS_CRITICAL_FULL T1_RU_JINR Site unknown in SAM3 here here The site contains one SRM in unknown status, which is propagated to the site availability     Green led     Understood &SRM
CMS CMS_CRITICAL_FULL T1_FR_CCIN2P3 Metric fails with different FQAN here here The metric JobSubmit is executed with two different FQANS. In SAM2, the results overlap each other. SAM3 considers the two metrics independent     Green led     Understood FQAN
CMS CMS_CRITICAL_FULL T1_TW_ASGC Service green, site red in SAM3 here here The services were in warning. It still needs some clarification why if the services were in warning, the availability gets reduced           under investigation  

T2 differences

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL All T2 Sites do not appear in SAM3 here here The sites do not belong to any group in the ALICE vofeed           Understood  
ATLAS ATLAS_CRITICAL GoeGrid Site available in SAM2, even if SRM not available here here The SRM instance is not taken by SAM2 to calculate the site availability. Probably, because the service is not in GOCDB           understood Unknown service ignored
ATLAS ATLAS_CRITICAL

H2_ATLAS_T2, IN2P3 _CC_T2,

LIP_COIMBRA, LRZ-MU, NO_NORDGRID_T2,ru-Moscow-SINP-LCG2, SE-SNIC-T2, TW_FTT, UKI-SOUTHGRID-RALPP, UNIBE-LHEP, UTA_SWT2, wuppertalprod

No Service --> Site green here here All the instances of the service are missing in SAM2. SAM2 computes the availability based on the other service. In SAM3, the service is in the vofeed, and the availability is properly calculated           understood New services in vofeed
ATLAS ATLAS_CRITICAL IN2P3_LAPP Metric fails for one FQAN here here The same metric can be executed with multiple FQAN. SAM2 considers the latest result received for the metric (independent of the FQAN). SAM3 considers the two metrics independent, and does an 'AND'           Understood FQAN
ATLAS ATLAS_CRITICAL prague-lcg2 Service ok -> Site RED in SAM3 here here There seems to be an SRM appearing in the vofeed and disappearing, dpmhead.ipv6, The status of this service is always red, causing the red in SAM3 availability           Understood New service in vofeed
ATLAS ATLAS_CRITICAL signet New site in SAM3   here The site appears in the vofeed           Understood New service in vofeed
ATLAS ATLAS_CRITICAL UKI-SCOTGRID-ECDF Additional block of 0% availability for ce6.glite.ecdf.ed.ac.uk which appears between 0900 and 1500 in the SAM3 page here here If we dig down into the metric history (here) we see that the metric is executed with two FQANs. SAM3 takes them as independent metric, SAM2 depends on the arrival time of each Tmetric           Understood FQAN
ATLAS ATLAS_CRITICAL UKI-SCOTGRID-DURHAM A block of time marked as missing for ce4.dur.scotgrid.ac.uk from 1200 to 1000 in SAM3 page here here A metric is supposed to be executed with multiple FQANS, and one of them is not being executed (here)           Understood FQAN
ATLAS ATLAS_CRITICAL UKI-SOUTHGRID-OX-HEP An extra service in the SAM3 interface, t2ce02.physics.ox.ac.uk, which is sitting at 0% here here The service appears in the ATLAS vofeed, so the service is included in SAM3           Understood New service in vofeed
ATLAS ATLAS_CRITICAL UKI-LT2-Brunel A block of time marked as missing for dc2-grid-66.brunel.ac.uk in SAM3 from 1500 to 1900 and for dc2-grid-26.brunel.ac.uk from 1800 to 1900. here here The metric is supposed to be executed with two different FQANS, and one of them is not runninghere           Understood FQAN
CMS CMS_CRITICAL_FULL T2_AT_Vienna Announced downtime does not apply to SRM here here The downtime had been announced properly in the GOCDB, and it was mapped to the CREAM-CE. However, it was not mapped to the SRM in SAM, since the service is called SRMv2 in the vofeed, and SRM in gocdb   Solved  
LHCb LHCB_CRITICAL All T2 Availability unknown for all T2 here here None of the T2 have an SRM. Since the algorithm is CREAM-CE * &SRMv2, all the sites are set to unknown. UPDATE: The logic of * has been modified in SAM3 to be 'and if data'           Understood Logic of 'and'

August T1 Summary Report

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL All No issues here here                
ATLAS ATLAS_CRITICAL NDGF-T1 Site Red here here The ARC-CE tests were not executed, and this propagated to the site availability   Agreed         Unknown service-->Site OK
  RAL-LCG2 Site Unknown     The ARC-CE was in an unknown state, and this propagated to the site availability   Agreed         Unknown service-->Site OK
  RRC-KI-T1 Site red     The site offers two SRM, and one of them is not passing the tests             All SRM
  Taiwan-LCG2 Site red     The site offers two SRM, and one of them is not passing the tests             All SRM
CMS CMS_CRITICAL T1_FR_CCIN2P3 Site red here here One of the metrics (JobSubmit) is executed with two different FQANS. SAM2 gets the status of the latest results, independent of the FQAN. SAM3 treats them as independent metrics             Multiple FQAN
    T1_RU_JINR 2 SRM instances     The site has two SRM instances, and one of them does not execute all the tests properly             All SRM
    T1_RU_JINR_Disk, T1_UK_RAL_Disk, T1_US_FNAL_Disk New sites in SAM3     The sites appear in the CMS vofeed. If they should not appear in the report, they should be removed from the vofeed             New site in VOfeed
LHCb LHCb_CRITICAL All No issues here here                

August T2 Report

Ignoring difference smaller than 3%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL     here here    
    KISTI, PACKGRID Sites do not appear in SAM3.     They are not in the vofeed  
ATLAS ATLAS_CRITICAL     here here    
    CA-MCGILL-CLUMEQ-T2 Downtime Rel 61%, Unk 1% Rel 53% Unk: 9%             to be investigated  
    UNIBE-LHEP ARC-CE Avl 63% Avl 9% Due to ARC-CE test not involved in ATLAS av/rel calculation by that time           Understood  
    praguelcg2 SRM profiles Avl 0% Avl 94% The site offers two SRM, and one of them is not passing the tests              
    wuppertal ARC-CE Avl 97% Avl 9% Due to ARC-CE test not involved in ATLAS av/rel calculation by that time           Understood  
    lrz-lmu ARC-CE Avl 100% Avl 15% Due to ARC-CE test not involved in ATLAS av/rel calculation by that time           Understood  
    GoeGrid SRM not in SAM2 Avl 96 % Avl 90%     SAM2/SAM3 give same results down to single tests outcome, not understood          
    SE-SNIC-T2 ARC-CE Avl 87 % Avl 2% Due to ARC-CE test not involved in ATLAS av/rel calculation by that time           Understood  
    WT2   Avl 96% Avl 84%     SAM2/SAM3 give same results down to single tests outcome, not understood          
    DESY-HH, DESY-ZN Site appears multiple times in SAM2, under different federations     The federation names are different per experiment. In SAM3, they appear only once,using the federation name of that experiment   ?          
CMS CMS_CRITICAL     here here    
    T2_IT_Pisa Downtime missing Avl 79% Avl 90%    
    T2_RU_ITEP Metric with multiple FQAN Avl 31% Avl 19%    
    T2_US_Florida Site appearing twice is SAM2        
    T2_US_Purdue Site appearing six times in SAM2        
    T2_US_Nebraska   Avl 92% Avl 99%    
    T2_TW_Taiwan Site appearing twice in SAM2 Avl 84% Avl 73%    
LHCb LHCb_CRITICAL     here here    
   

LCG.DESYHH.DE, DESYZZ.DE,LCG.DESY

SAM2 presents DESY.DE, SAM3 presents the other 2 sites        
    LAL.fr, LLR.fr, LPNHE.fr Federation does not appear in SAM3        
    ARC_BRISTOL.UK, ARC_OXFORD.UK, DURHAM.UK Sites appearing in SAM3 (they are in the vofeed)        

September T1 Summary Report

Note On the 25th of September, the database of SAM3 was moved from integration to production. That explains the grey day. We will work on recovering that data

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL CCIN2P3 Downtime missing in SAM2 here here
    SARA Downtime missing in SAM2 here here
ATLAS ATLAS_CRITICAL CERN-PROD Two srm instances here here
    FZK Two srm instances
    CCIN2P3 Downtime missing in SAM2
    NDGF ARC-CE
    RRC-K1 Two srm instances
    SARA Missing downtime in SAM2
Taiwan-LCG2 Missing downtime in SAM2 and two srm instances
CMS CMS_CRITICAL T1_CH_CERN Removing site from vofeed here here
    T1_RU_JINR Two srm instances
    CCIN2P3 Downtime missing in SAM2
LHCb LHCb_CRITICAL CCIN2P3 Downtime missing in SAM2 here here

September T2 Report

Ignoring difference smaller than 3%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL     here here    
    Prague   92% 87%
    Bari   35% 86% The site has several CREAM-CE. At least one of them was working during most of the time. The SAM3 calculation is more accurate
    Trigrid-Catania, PAKGRID New sites in SAM3
    KISTI   100% 85% The service was in an unknown state during a downtime. The downtime is not visible in SAM2. SAM3 data is more accurate
    SauPaulo   99% 76% The service was in an unknown state during a downtime. The downtime is not visible in SAM2. SAM3 data is more accurate
ATLAS ATLAS_CRITICAL     here here    
    Prague Two srm instances 97% 1%
UNIBE ARC-CE
DESY-ZN, DESY-HH Sites appear in SAM2 under multiple federations
wuppertal, LRZ-LMU, SE-SNIC-T2 ARC-CE
IN2P3-LAPP Probe run with two credentials, and one of them fails. 90% 81%
DURHAM ARC-CE 100% 91%
CMS CMS_CRITICAL     here here    
T2_CE_UCL   91% 86%
T2_ES_IFCA   97% 85%
T2_RU_ITEP   72% 66%
T2_BR_UERJ 7% 42%
T2_TW_Taiwan   90% 80%
T2_UA_KIPT   85% 75%
T2_DE_DESY Site appearing in multiple federations in SAM2
T2_US_UCSD Missing site in SAM3
T2_US_Florida, T2_US_Purdue Site appearing multiple times in SAM2
LHCb LHCb_CRITICAL     here here    
LCG.Durham.uk, ARC.Bristol.uk, ARC.Oxford.uk New empty sites appearing in SAM3

October T1 Summary Report

Note Ignoring differences smaller than 1%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL CERN Downtime in SAM2 and maintenance in SAM3 here here
    RRC_KI_T1 Availability in SAM2 and SAM3 differs in 6 %(!); there is 2 days “unknown” instead maintenance in SAM3 reliability plot here here Availability is calculated not correctly in SAM3: 92% is a correct result (as in SAM2)
ATLAS ATLAS_CRITICAL CERN-PROD Downtime is missing in SAM2; availability and reliability in SAM2 and SAM3 differ by3 % here here There are several CREAM-CEs at CERN. One of them is always up, so it has to be investigated why the CREAM-CE Availability in SAM3 is not green
    FZK More Downtime periods in SAM3 than in SAM2; availability and reliability in SAM2 and SAM3 differ by2 %     There are several CREAM-CEs at FZK and only one of them was in down state on 04-05 October; there is no information about down time at any days after October 5 in SAM3 Service availability historical plots but there are several days with down time in SAM3 summary plot
    NDGF A part of Downtime and Maintenance is missing in SAM2; availability and reliability in SAM2 and SAM3 differ by ~3 %     SAM3 results seem to be more correct according SAM3 Service availability historical plots
    NIKHEF Maintenance in SAM2 instead of Downtime in SAM3; availability and reliability in SAM2 and SAM3 differ by ~8 %     SAM3 results seem to be correct according SAM3 Service availability historical plots
    RRC-KI-T1 Missing downtime in SAM2; availability and reliability in SAM2 and SAM3 differ by 81 % and 89% rel-ly     The site has several SRM instances. Since the algorithm request 'all SRM', and one of the SRMs is down, the availability goes down. SAM3 seems to be correct
Taiwan-LCG2 Missing downtime in SAM2; availability and reliability in SAM2 and SAM3 differ by ~4 %     SAM3 result corresponds SAM3 Service availability historical plot
CMS CMS_CRITICAL T1_CH_CERN Still not removed in SAM2 here here
    T0_CH_CERN A part of downtime missing in SAM2; availability and reliability in SAM2 and SAM3 differ by2 %     SAM2 and SAM3 availability results differ on 22 and 23 October and it is seen from SAM3 Service availability historical plot that availability was defined not correctly in SAM3 on that days
    T1_DE_KIT No Downtime in SAM2; availability and reliability in SAM2 and SAM3 differ by 3.5 %     All SRMv2 OK; a lot of CREAM-CEs also OK - but on 06 and 07 October there is downtime in SAM3 summary report - it should be investigated
    T1_TW_ASGC The plots are the same in SAM2 and SAM3 but availability and reliability differ by 2.7 %     SAM3 result is correct
LHCb LHCb_CRITICAL LCG-RRCKI.ru No information on this site in SAM2 here here

October T2 Report

Ignoring difference smaller than 3%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL     here here    
Kolkata   98% 95% Availability of 95% corresponds SAM3 Service availability historical plots
  Trigrid-Catania Site in SAM3 and no such a site in SUM2
  SaoPaolo   79/94% 52/91% Availability of 52% corresponds SAM3 Service availability historical plots
  SUT The percentage of “unknown” state is 93% in SAM3 and 19% in SAM2     No data for 1-28 October in SAM3 Service availability historical plots
ISMA   92% 84% Availability equals 87% according SAM3 Service availability historical plots
KNU   0% N/A Availability should be 0% according SAM3 Service availability historical plots
KNU The percentage of “unknown” state is 26% in SAM3 and 0% in SAM2     6 days of "unknown" state in SAM3 Service availability historical plots - it is 19% (not 26%)
ATLAS ATLAS_CRITICAL     here here    
CSCS-LCG2 The percentage of “unknown” state is 12% in SAM3 and 0% in SAM2     2 days (~10%) are in "unknown" state SAM3 Service availability historical plots
    praguelcg2   54% N/A Availability should be equal to 0% in SAM3
GoeGrid   100% 91% Availability of 91% corresponds SAM3 Service availability historical plots
GoeGrid The percentage of “unknown” state is 4% in SAM3 and 0% in SAM2     There are only 8 hours of "unknown" state (on October 1) as it is seen in SAM3 Service availability historical plots and it is ~ 1% but not 4%
DESY-ZN, DESY-HH Sites still appear in SAM2 under multiple federations
wuppertal   95% 69% Availability of 69% corresponds SAM3 Service availability historical plots
wuppertal The percentage of “unknown” state is 31% in SAM3 and 37% in SAM2     The percentage of “unknown” state as 31% corresponds SAM3 Service availability historical plots
IN2P3-LAPP   99% 96% Availability of 96% corresponds SAM3 Service availability historical plots
CYFRONET-LCG2   99% 94% Availability of 78% corresponds SAM3 Service availability historical plots
PSNC   78% 66% Availability of 66% corresponds SAM3 Service availability historical plots
PSNC The percentage of “unknown” state is 12% in SAM3 and 5% in SAM2     According SAM3 Service availability historical plots, the percentage of “unknown” state is about 9%
RRC-KI The percentage of “unknown” state is 12% in SAM3 and 15% in SAM2     SAM3 result (12%) fully corresponds SAM3 Service availability historical plots
SE-SNC-T2   96% 91% The availability of 91% corresponds SAM3 Service availability historical plots
TR-10-ULAKBIM   79% 76% The availability of 76% corresponds SAM3 Service availability historical plots
UKI-SOUTHGRID-BHAM –HEP   86% 77% The availability of 77% corresponds SAM3 Service availability historical plots
HU-ATLAS-Tier2   76% 72% If 18 days in "unknown" state are not taken into account, the availability of 72% is correct
OU_OCHEP_SWT2   42% 63% The availability of 63% corresponds SAM3 Service availability historical plots
CMS CMS_CRITICAL     here here    
T2_BE_UCL   90% 71% The availability of 71% corresponds SAM3 Service availability historical plots
T2_IN_TIFR   87% 74% The availability of 74% corresponds SAM3 Service availability historical plots
T2_IN_TIFR The percentage of “unknown” state is 45% in SAM3 and 32% in SAM2     45% is a correct result in accordance with SAM3 Service availability historical plots (with "Hourly" granularity)
T2_RU_SINP   0% N/A Currently T2_RU_SINP is not in operation state
T2_BR_UERJ   86% 89% The availability of 89% corresponds SAM3 Service availability historical plots
T2_BR_UERJ The percentage of “unknown” state is 5% in SAM3 and 2% in SAM2     5% is a correct result
T2_US_Florida   91% 87% The availability of 87% corresponds SAM3 Service availability historical plots
T2_US_Nebraska   98% 64% The availability of 64% corresponds SAM3 Service availability historical plots
T2_US_UCSD   99% 77% The availability of 77% corresponds SAM3 Service availability historical plots
T2_US_UCSD The percentage of “unknown” state is 50`% in SAM3 and 0% in SAM2     "Hourly" granularity in SAM3 Service availability historical plots shows exactly 50% as "unknown" state but it is not clear why for 4 days of that period the availability equals "0" and for 1 day - "0.56"
T2_TH_CUNSTDA   95% 92% The availability of 92% corresponds SAM3 Service availability historical plots
T2_TW_Taiwan   90% 87% The availability of 87% corresponds SAM3 Service availability historical plots
T2_DE_DESY Site appearing in multiple federations in SAM2
T2_US_Florida, T2_US_Purdue, T2_TW_Taiwan Sites appearing multiple times in SAM2
LHCb LHCb_CRITICAL     here here    
DESYZN.de, DESYHH.de Sites appearing multiple times in SAM2
LCG.Durham.uk, ARC.Bristol.uk, ARC.Oxford.uk Empty sites in SAM3
LCG-RAL-HEP.uk New empty site appearing in SAM3

November T1 Summary Report

Note Ignoring differences smaller than 1%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL No issues   here here
ATLAS ATLAS_CRITICAL CERN-PROD Downtime is missing in SAM2; availability and reliability in SAM2 and SAM3 differ by 2% and 1% here here It's seems to be a mistake in SAM3: there was NO downtime for CERN-PROD
    BNL-ATLAS Availability and reliability in SAM2 and SAM3 differ by 1%     Availability in SAM3=99.7% (100% in Summary plot)
    FZK More Downtime periods in SAM3 than in SAM2; availability and reliability in SAM2 and SAM3 differ by 10 % !     SAM3 results in Summary plot differ from the FZK Service Availability historical plots - it should be investigated
    IN2P3 Availability and reliability in SAM2 and SAM3 differ by 1 %     Availability in SAM3 is 99.7% (100% at Summary plot)
    INFN Availability and reliability in SAM2 and SAM3 differ by 2 % and 1%     Availability in SAM3 on November 18 is calculated not correctly (see Service Availability historical plots) - it should be 86% instead 100% - then Availability for November will be not 100% but 99%
    RRC-KI-T1 Missing downtime in SAM2; availability and reliability in SAM2 and SAM3 differ by 35 % and 34% rel-ly     The results in SAM3 seem to be correct
    Taiwan-LCG2 Missing downtime in SAM2; availability and reliability in SAM2 and SAM3 differ by 9 %     The results in SAM3 seem to be correct
    pic Availability in SAM2 and SAM3 differs by 1 %   Availability in SAM3 on November 18 and 26 are calculated not correctly (see Service Availability historical plots) - it should be 88% and 86% instead of 100% - then Availability for November will be not 100% but 99%
CMS CMS_CRITICAL T1_CH_CERN Still not removed in SAM2 here here
    T1_ES_PIC Availability in SAM2 and SAM3 differs by 1 %     SAM3 results seem to be not correct (Nov.19 and 27): see
    T1_FR_CCIN2P3 Availability and reliability in SAM2 and SAM3 differ by 2%     SAM3 results seem to be not correct (Nov.18 and 19): see
    T1_TW_ASGC The plots are the same in SAM2 and SAM3 but availability and reliability differ by 3 %     SAM3 result is not correct; moreover, on Nov.19 availability is calculated wrong: 100% instead of 98%: see
LHCb LHCb_CRITICAL PIC.es No maintenance time in SAM2; availability in SAM2 and SAM3 differs by 1% here here SAM3 result seems to be correct

November T2 Report

Ignoring difference smaller than 3%

VO Profile Sites Issue SAM2 link SAM3 link Explanation Agreed by ALICE Agreed by ATLAS Agreed by CMS Agreed by LHCb Agreed by OPS Status Category
ALICE ALICE_CRITICAL     here here    
GRIF_IRFU   75% 67% SAM3 result (67%) is correct)
Catania   85% 89% SAM3 result is not correct - it should be 85%
  Trigrid-Catania Site in SAM3 and no such a site in SUM2
  Bratislava   92% 96% SAM3 result is correct
  BITP   0% / 0% N/A / N/A SAM3 result is 0% (not "N/A") - see
KNU The percentage of “unknown” state is 100% in SAM3 and N/A in SAM2     SAM3 result is correct
ATLAS ATLAS_CRITICAL     here here    
    praguelcg2   81% / 81% N/A / N/A SAM3 result is 0% (not "N/A") - see
DESY-ZN, DESY-HH Sites still appear in SAM2 under multiple federations
UNI-FREIBURG   79%/79% 64%/64% SAM3 historical plot shows 0% of availability
UNI-FREIBURG The percentage of “unknown” state is 49% in SAM3 and 44% in SAM2     The percentage of “unknown” state in SAM3 is 60%
MPPMU The percentage of “unknown” state is 0% in SAM3 and 4% in SAM2     SAM3 result is correct
GRIF-IRFU   65%/85% 57%/63% SAM3 result seems to be not correct; SAM3 result for avail. on November 18 also is not correct see
PSNC The percentage of “unknown” state is 14% in SAM3 and 7% in SAM2     SAM3 result seems to be not correct - see; in particular, SAM3 result for Nov.29 is not correct
RO-14-ITIM   93% 84% SAM3 hist.plot - see results in 76% of aviail.; moreover, see here that result fot Nov.18 is not correct
ru-PNPI   93% / 93% 88% / 88% 88% is correct but see that SAM3 result for Nov.7 is not correct
FMPhi-UNIBA   92% 88% SAM3 result should be equal 81%; by the way, there is extra and wrong name "FMPhi-UNIB" in the "Sites" menu in SAM3
TR-10-ULAKBIM   75%/75% 69%/60% SAM3 result is ~64% see
UKI-SCOTGRID-ECDF The percentage of “unknown” state is 6% in SAM3 and 2% in SAM2     Only 3 hours of "unknown" in SAM3 on Nov.23 see
UKI-SCOTGRID-GLASGOW   97%/97% 93%/93% SAM3 result is 93% but SAM3 result for Nov.9 is not correct - if it will be corrected, SAM3 result will be also 97% as in SAM2 see
UKI-SOUTHGRID-BHAM-HEP   96%/96% 93%/93% The same incorrectness in SAM3 as for UKI-SCOTGRID-GLASGOW see Nov.23 here
MWT2   95% 98% 98% is a correct SAM3 result but there is incorrect result in SAM3 for Nov.18 see
HU_ATLAS_Tier2   0%/0% N/A / N/A There are 2 days of avail.=0% and 28 days with "unknown" status in SAM3 - see
CMS CMS_CRITICAL     here here    
T2_AT_Vienna   83%/92% 78%/88% 78% is correct in SAM3 but there is incorrectness onSAM3 result for Nov.19 - see
T2_AT_Vienna The percentage of “unknown” state is 5% in SAM3 and 10% in SAM2     17 hours of "unknown" and more 17 hours of "unknown" - it is ~5%
T2_BE_IIHE   95% (Rel.) 85% (Rel.) SAM3 result seems to be not correct
T2_BE_UCL   96%/98% 83%/83% SAM3 result seems to be correct
T2_BE_UCL The percentage of “unknown” state is 16% in SAM3 and 19% in SAM2     SAM3 result seems to be correct - see
T2_CN_Beijing   100%/100% 97%/97% 97% is a strange result: it could be obtained in a case of CEs avail. on Nov.19 and Nov.are taken into account - but they aren't... - see
T2_DE_RWTH   93%(Rel.) 98%(Rel.) SAM3 result seems to be not correct - see
T2_ES_CIEMAT   99%(Rel.) 82%(Rel.) SAM3 result seems to be not correct - see
T2_ES_IFCA   99%/99% 92%/92% SAM3 result seems to be correct
T2_FR_GRIF_IRFU   65%/85% 68%/68% SAM3 result seems to be not correct - see
T2_FR_CCIN2P3   98%/98% 82%82% SAM3 result seems to be correct for exception Nov.18 - see
T2_FR_CCIN2P3 The percentage of “unknown” state is 20% in SAM3 and 30% in SAM2     20% seems to be correct but a strange result for "unknown" period on Nov.198- see
T2_GR_Ioannina   0%/0% N/A / N/A SAM3 also shows 0%
T2_IN_TIFR   93%/93% 80%/80% 80% seems to be correct but incorrect results for Nov.20 and Nov.22
T2_IN_TIFR The percentage of “unknown” state is 21% in SAM3 and 6% in SAM2     SAM3 result seems to be correct
T2_PT_NCG_LIsbon   98%/98% 92%/92% 92% is correct for SAM3 resulting plot but there is incorretness for Nov.19 - see
T2_RU_RRC_KI   0%/0% N/A / N/A SAM3 alos shows 0%
T2_RU_SINP   0%/0% N/A / N/A SAM3 alos shows 0%
T2_DE_DESY Site appearing in multiple federations in SAM2
T2_US_Florida, T2_US_Purdue, T2_TW_Taiwan Sites appearing multiple times in SAM2
T2_US_Nebraska   100%/100% 80%/80% SAM3 result seems to be correct
T2_US_UCSD The percentage of “unknown” state is 1% in SAM3 and 4% in SAM2     1% is correct in the sense of SAM3 resulting plot but there are strange results for "unknown" period on Nov.19 and Nov.12 and 13
LHCb LHCb_CRITICAL     here here    
DESYZN.de, DESYHH.de Sites appearing multiple times in SAM2
LLR.fr   98% 75% SAM3 result is not correct: it should be equal 88%
LLR.fr The percentage of “unknown” state is 12% in SAM3 and 27% in SAM2     The percentage of "unknown" state in SAM3 is 23%
PNPI.ru   99%/99% 95%/95% SAM3 result seems to be correct
LCG.Durham.uk The percentage of “unknown” state is 11% in SAM3 and 0% in SAM2     If "missing" periods are considered as "unknown", SAM3 result is correct
ARC.Bristol.uk   0%/0% N/A / N/A As it is seen from, SAM3 result is also 0%
ARC.Oxford.uk The percentage of “unknown” state is 11% in SAM3 and 0% in SAM2     If "missing" periods are considered as "unknown", SAM3 result is correct
RAL-HEP.uk The percentage of “unknown” state is 11% in SAM3 and 0% in SAM2     If "missing" periods are considered as "unknown", SAM3 result is correct

-- PabloSaiz - 18 Aug 2014

Edit | Attach | Watch | Print version | History: r46 < r45 < r44 < r43 < r42 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r46 - 2014-12-12 - ElenaTikhonenko
 
    • 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 &© 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback