IPFS file replication on all nodes in Private network

1

IPFS have any replication factor, so that we can be sure that all files are available on all the nodes and we don’t have a single point of failure.

Eg if we have setup an private network with 3 Org. Org1 uploaded a file and shutdowns.Dose other node will have copy of the uploaded file if they have not asked for it before Org1 going off?

2

If neither node 2 nor node 3 requested the file added to node 1, and node 1 shuts down, then nodes 2 & 3 don’t have it. In this case the nodes have to pin the file beforehand (manual node job), or you automate it with ipfs-cluster:

https://dist.ipfs.io/#ipfs-cluster-ctl

3

Hey @JayBrown,
Thanks a lot for the response… Will try to use the cluster-ctl

4

Hi @JayBrown, i created a cluster on 1 server and pined the 2 nodes.
I am facing the below issue when node 2 or node 3 requested the file added to node 1.

17:15:59.237 ERROR   ipfshttp: error getting:Get http://xx.xxx.xx.xxx:5001/api/v0/repo/stat: dial tcp xx.xxx.xx.xxx:5001: getsockopt: connection refused ipfshttp.go:704
17:15:59.237 ERROR   ipfshttp: Get http://xx.xxx.xx.xxx:5001/api/v0/repo/stat: dial tcp xx.xxx.xx.xxx:5001: getsockopt: connection refused ipfshttp.go:821
17:15:59.237 ERROR  p2p-gorpc: Get http://xx.xxx.xx.xxx:5001/api/v0/repo/stat: dial tcp xx.xxx.xx.xxx:5001: getsockopt: connection refused client.go:125
17:15:59.237 ERROR   diskinfo: Get http://xx.xxx.xx.xxx:5001/api/v0/repo/stat: dial tcp xx.xxx.xx.xxx:5001: getsockopt: connection refused disk.go:101
5

Can’t say. Anyone having experience with cluster-ctl?

6

@JayBrown what should be the minimum system configuration for running the IPFS-cluster ?

7

As far as I can see, there’s always a 386 binary, too, so I assume that all IPFS releases including cluster-ctl have very generous minimum system requirements. What I don’t see is a PowerPC binary. :wink: