Add a minimally useful incident "handling" to maintd: collect incidents regularly and expose them at a central place. Also add a command to clear (manually) handled incidents.
Klaus Aehlig (6): Also for incidents, call their UUID "uuid" Extend MaintenanceData by incidents Also keep the list of incidents in memory Expose the current list of incidents Collect diagnose data from MonDs Add a command to unconditionally forget repair events Makefile.am | 1 + doc/design-repaird.rst | 2 +- lib/client/gnt_cluster.py | 19 +++++ lib/objects.py | 2 +- man/ganeti-maintd.rst | 36 +++++++++ man/gnt-cluster.rst | 10 +++ src/Ganeti/MaintD/CollectIncidents.hs | 129 +++++++++++++++++++++++++++++++ src/Ganeti/MaintD/MemoryState.hs | 34 +++++++- src/Ganeti/MaintD/Server.hs | 3 + src/Ganeti/Objects/Maintenance.hs | 51 ++++++++++++ src/Ganeti/WConfd/ConfigModifications.hs | 15 ++++ src/Ganeti/WConfd/Core.hs | 6 ++ test/hs/Test/Ganeti/Objects.hs | 19 +++++ 13 files changed, 323 insertions(+), 4 deletions(-) create mode 100644 src/Ganeti/MaintD/CollectIncidents.hs -- 2.5.0.rc2.392.g76e840b
