Keep our IPFS nodes connected to each other #44

New Issue

greg · 2019-04-26T09:17:54Z

greg commented

2019-04-26 09:17:54 +00:00

dev and andromeda keep disconnecting from each other, both after initial startup (they're in the list of bootstrap nodes for each other) and after manual connect

greg self-assigned this 2019-04-26 09:17:58 +00:00

greg changed title from ~~Keep our IPFS connected to each other~~ to Keep our IPFS nodes connected to each other

2019-04-26 09:18:56 +00:00

greg commented

2019-04-26 10:27:31 +00:00

After seeing an error in the ipfs-cluster logs on dev (IPFS unsuccessful: 500: open /home/ipfs/.ipfs/datastore: too many open files), I have raised the maximum number of open files (manually for now) by adding ``LimitNOFILE=64000` to the systemd unit for dev and andromeda, this appears to have fixed the issue. dev and andromeda have been in each other's swarm peers for almost an hour now

After seeing an error in the ipfs-cluster logs on dev (`IPFS unsuccessful: 500: open /home/ipfs/.ipfs/datastore: too many open files`), I have raised the maximum number of open files (manually for now) by adding ``LimitNOFILE=64000` to the systemd unit for dev and andromeda, this appears to have fixed the issue. dev and andromeda have been in each other's swarm peers for almost an hour now

raucao commented

2019-04-26 10:55:53 +00:00

Hooray! Good catch.

greg referenced this issue from a commit

2019-04-26 14:00:20 +00:00

Update the ipfs cookbook to set a maximum number of files The default is too low and was treated like a low disk space alert by ipfs and ipfs-cluster, set it to 64000 Closes #44

raucao commented

2019-04-26 14:40:35 +00:00

Confirmed to be solved by checking a new contribution on kredits-web just now.

raucao closed this issue

2019-04-26 14:41:06 +00:00

Sign in to join this conversation.