Get a large file and watch bitswap acting.
Reset the bitswap statistics.
ipfs bitswap stat
ipfs shutdown
ipfs daemon &
ipfs bitswap stat
Watch the statistics in one terminal.
while true
do
ipfs bitswap stat | grep -v Qm | grep -v 12D
sleep 1
done
Watch the bitswap ledger in one terminal.
while true
do
for peer in $(ipfs swarm peers | awk -F/ '{ print $NF }')
do
ledgerdata=$(ipfs bitswap ledger $peer)
if [[ -n "$(echo $ledgerdata | grep -v '\t0$' | grep -v '\t0.000000$' | grep Bytes)" ]]
then
date
echo $peer
echo $ledgerdata
echo
fi
done
sleep 1
done
Watch the number of connected peers in one terminal.
while true
do
printf "\rNumber of connected peers: %s" $(ipfs swarm peers | wc -l)
sleep 1
done
Start the download in one terminal.
ipfs get -o /tmp/bigfile QmYi6yy3ygc3dNJETqFvvyXdZx28dNyNbZLHkiimhd8Jzd
Note: Do you have it already in your repo? You can remove it, i.e. ipfs repo gc.
If the download is stalling, you can do the same download from another node in parallel.
ToDo: What is different, if a peer nearby has the big file already downloaded?