I’m trying to build up an IPFS Cluster of 6 node that is running on top of IPFS public node.
I’m using ipfs-cluster to manage the pinning replication. In my setup I put 3 pin for the replication factor.
All 6 nodes are running in the same local network.
The cluster is working fine, pinned content are correctly pinned over 3 other nodes. But I’m not able to get the content from the IPFS gateway where the content is note pinned. Is it normal ?
I would try to retrieve the content through a different gateway to see if that works. You could use the dweb gateway or the web3.storage gateway too to try and get your data retrieved faster. Gateways have access to public ipfs content as i’m aware but its not always easy for a Gateway to retrieve that content through the network quickly especially if your ipfs node(s) are all on the same LAN(might be many hops from the nearest gateway server)
It might be better to send you content to a pinning service that has gateway onramps like Web3.Storage. Their gateway is pretty fast as are a few others.
I do not know if this is a solution to your problem but a way to troubleshoot to see if that particular gateway is the issue.
That’s also a good test, if you can’t find a public gateway that can load the CID, it’s either a content publishing or a network reachability problem of your nodes.
You can test content publishing/routing problems using pl-diagnose and seeing if your CIDs are on the DHT. If they’re not on the DHT, it’s likely a content publishing problem.
How many files are you pinning approximately?
I’d recommend reading this blog post to learn more about how this works:
Specifically, I’d consider updating the reprovider strategy to either roots or pinned, which could decrease the number of CIDs needs to publish and the amount of time it spends publishing