Radim Kolar
Radim Kolar
We know what we need to download (cache size and player head position is known). Do intersection of "download plan" and "what is cached" - Do not release from cache...
Accelerated DHT Client is unuseable, it creates thousands of threads effectively stopping all other workloads on machine while working - it is doing 10 minutes long batches.
450k objects
I dont think this is a bug. It maybe confusing and need some UI adjust but behavior is correct.
This Can be done by adding reference counter property to CID
dont auto trigger gc not only its quite long and io heavy slowing machine, it will delete too much data. I like current behavior. If we are bellow GC size...
This will not be fast too. you need first to build recrusive list of all cids from pin roots and then start travrl your cid to be deleted and delete...
48 hours is too much, 3 hours is more realistic since 80% of IPFS nodes in my scan have uptime lower then 2h. They are probably Brave run nodes.
changed IP is not a problem, DHT is organised by Peer Id. It will quickly switch to new node IP.
I see people are setting DHT to client mode manually to reduce CPU and Bandwidth load on their servers. I have several servers and only one is set to dhtserver...