Known HLT operational issues during Run II

Known ERROR/WARINING messages which can be ignored

• LoKiSvc.REPORT: LoKi::Particles::CosineDirectionAngleWithTheBestPV: VertexHolder base is invalid, return -3000 [Hlt2CharmHadD02HH_D0ToPimPip_LTUNBCombiner] StatusCode=FAILURE

• Hlt2CharmHadD02HH_D0ToKmPipCombiner: CombineParticles:: The WARNING message is suppressed : 'PV re-fit failed' StatusCode=FAILURE

• Hlt2CharmHadD02HH_D0ToKmPip_LTUNBCombiner: CombineParticles:: The WARNING message is suppressed : 'PV re-fit failed' StatusCode=FAILURE

• Hlt2TrackEff_D0ToKpiPionProbeDstDTFFilter: DecayTreeFitter::Fitter: Warning: Error matrix not positive definite. Changing status to failed.

• Hlt2TrackEffDownstream1N3BodyCombiner.MomentumCombiner: MomentumCombiner:: Only 4-momenta sum will be used (no ParticleTransporter specified) [Hlt2TrackEffDownstream1N3BodyCombiner] StatusCode=FAILURE

• ToolSvc.LoKi::VertexFitter: LoKi::VertexFitter:: Error from Kalman-step, skip [Hlt2ConvPhotonLL] StatusCode=405

• ToolSvc.LoKi::VertexFitter: LoKi::VertexFitter:: _transport(): the error from transport(), ignore [Hlt2CharmHadSpec_DpPi0TurboSpec_DpPi0Combiner] StatusCode=412

• Hlt2TauPiPi0CombinerCombiner.MomentumCombiner: MomentumCombiner:: Only 4-momenta sum will be used (no ParticleTransporter specified) [Hlt2TauPiPi0CombinerCombiner] StatusCode=FAILURE

• Hlt2TrackEff_D0ToKpiKaonProbeDstDTFFilter.LoKi::DecayTreeFit: LoKi::DecayTreeFit:: The WARNING message is suppressed : 'Fitter failed status' StatusCode=111

Appeared during testing of 5 TeV reference run.

• Hlt2SelReportsMaker: HltSelReportsMaker:: Only some particles in the trigger selection Hlt2LowMultPi0Decision have PV association - none stored
• Hlt2SelReportsMaker: HltSelReportsMaker:: PV selection _ found via Particle2Vertex for Hlt2LowMultPi0Decision has unregistered name - no associations stored

Appeared during PbPb run

• Hlt2VeloMuonTrackingProbeProtoPAlg: ChargedProtoParticleMaker:: No Tracks at 'Hlt2/TrackEff/Track/VeloMuon'

StdInfoSubBank too large to store. No Std Info will be saved. [FIXED SINCE JULY TS 2015]

• Some events in HLT2 create too many candidates which corrupts the RawBank and causes part of the trigger information about the event (but not the event itself) to be lost. See http://lblogbook.cern.ch/HLT+Trigger/761 . This error can be ignored.

HLT1 JobOptions Writer goes into error

• This can be caused by having too few subfarms available. The recipe requires that at least 61 sub farms are included. So it this problem occurs, try and re-add excluded farms.

HLT1 Rate "dial" on shift leader's computer stops updating

• This sometimes happens - check that the rate on the monitoring plots is still reasonable.

HLT2 Monitoring histograms can only be viewed in history mode

• When looking in online mode, HLT2 histograms are always empty. This is a known "feature", we are working on a fix. For now they must be viewed in history mode.

-- VladimirGligorov - 2015-07-12