Warm up rpki-client cache 30 min before scheduled run #25
Conversation
|
Would it make sense to fetch the repository data regularly and maintain the latest set in the asmap-data repo? |
No, I don't think that makes sense. First of all, we would need to pay for the storage space we would need for that. And I also don't see any upside to it. The whole RPKI repository infrastructure is optimized for (decentralized) data availability so if there is a reason why we can't get the data from it reliably we should first find out what the issue is. GitHub is also not a very reliable service IMO so we would need to think about redundancy etc. and we would also make RPKI a centralized data source because if something goes wrong only our data might be used. I don't want to risk that. |
This is an idea to improve the issue described in #19. We warm up the rpki-client cache once 30min before the scheduled run.
Keeping this as a draft for now. I need to think about if we want to do some more refactorings, if this is really helping and if it is stable. I don't think we need to have this soon. It would be good enough for the next runs in the short-term that we have debug logging and see if the issue #19 comes up more regularly. If it comes up more regularly maybe we see some other issues in those logs that give us a better idea to actually fix the root cause and this might turn out to not be needed at all.