Pack relations tables in Tesla

Always pack relations tables, as subsequent processing steps can 'corrupt' unpacked tables. See LHCBPS-1810 for more details.

Please propagate to 2018-patches, run2-patches, and master.

@erodrigu it would be helpful if this could be quickly released in a DaVinci/2017-patches release, so we can perform a final validation of the simulation workflow for 2017 ASAP. @kreps @gcorti FYI.

the list is just https://gitlab.cern.ch/lhcb/Hlt/blob/fffc5e442e124424fa73871807848a80ebcb0345/Hlt/HltConf/python/HltConf/HltPersistReco.py#L119-136 + 'pPhys/P2MCPRelations', 'pPhys/PP2MCPRelations', right?

That's right! I wondered if this list existed elsewhere already..

Unfortunately I don't think we can re-use those definitions as they're in Hlt.

But it does make sense to state what the list is, as Rosen points out, and refer to the relevant Hlt configurable.

yeah, a comment won't hurt.

Is

List of locations produced by PackParticlesAndVertices

not sufficient? It's actually HltPersistReco that doesn't explain what those lists are

approved this merge request

• [2018-11-27 00:13] Validation started with lhcb-2017-patches#583

merged

mentioned in commit bd12dabb

mentioned in commit dbfbd86a

mentioned in merge request !273 (merged)

mentioned in commit f6f5d851

mentioned in merge request !274 (merged)

mentioned in commit ad91c862

mentioned in merge request !275 (merged)