So we have a service has defined some Python callbacks. They work fine, except that during an NSO restart that included a package revision upgrade we encountered this:
<CRIT> 11-Jan-2021::21:00:03.007 nso-netlab ncs[105]: - no registration found for callpoint routing-domain-bgp-dg-validate/validate of type=external
<ERR> 11-Jan-2021::21:00:03.007 nso-netlab ncs[105]: devel-c no registration found for callpoint routing-domain-bgp-dg-validate/validate of type=external path /ncs:services/routing-domain{ESNET}/bgp
<CRIT> 11-Jan-2021::21:00:03.025 nso-netlab ncs[105]: - CDB: Upgrade failed: Upgrade transaction failed to validate: application communication failure
Hint: Try using staged start using ConfD start-phases or the --ignore-initial-validation option
"CDB: Upgrade failed: Upgrade transaction failed to validate: application communication failure\nHint: Try using staged start using ConfD start-phases or the --ignore-initial-validation option\n"I am sure I do not fully understand the NCS application lifecycle but from what I've found, I think that my validation code lives in a Python VM that only starts in phase2... but the CDB upgrade in phase0 attempts to validate anyway.
Using the --ignore-initial-validation workaround does allow NCS to start but I would rather not do that. This is with NSO 5.3.1.
Looking for input on how to do this the right way.
Many thanks!
