Stake pool setup errors - cardano-cli version -bash: cardano-cli: command not found

Then I attempted to query and this is the result

cardano-cli query tip --mainnet
Command failed: query tip Error: Error while looking up environment variable: CARDANO_NODE_SOCKET_PATH Error: “CARDANO_NODE_SOCKET_PATH”

Have you run these commands:

echo export CARDANO_NODE_SOCKET_PATH="${CNODE_HOME}/sockets/node0.socket" >> $HOME/.bashrc
source $HOME/.bashrc

If yes, can you run: ls $CARDANO_NODE_SOCKET_PATH

If there’s some error please run: ps auxf | grep -i "cardano-node" and tell me what path is configured for the --socket-path parameter.

Thanks ok heres the results still no Socket which is strange as when I run a node on my laptop its fine.

ps auxf | grep -i “cardano-node”
matt 872 65.0 1.6 1074350820 131004 ? Dl 13:18 3:05 _ /home/matt/.cabal/bin/cardano-node run --topology /opt/cardano/cnode/files/topology.json --config /opt/cardano/cnode/files/config.json --database-path /opt/cardano/cnode/db --socket-path /opt/cardano/cnode/sockets/node0.socket --port 6333 --host-addr 0.0.0.0
matt 7544 0.0 0.0 4628 888 pts/0 R+ 13:22 0:00 _ grep -i cardano-node

ls $CARDANO_NODE_SOCKET_PATH
ls: cannot access ‘/opt/cardano/cnode/sockets/node0.socket’: No such file or directory

echo $CARDANO_NODE_SOCKET_PATH
/opt/cardano/cnode/sockets/node0.socket

export CARDANO_NODE_SOCKET_PATH="${CNODE_HOME}/sockets/node0.socket" >> $HOME/.bashrc

echo $CARDANO_NODE_SOCKET_PATH
/opt/cardano/cnode/sockets/node0.socket

How much RAM do you have? Please type: free -m

I am wondering if its something to do with Firewall ?
I set a custom port on the Node during config, I have just added a rule for this now

Anyway here is the memory

8GB

ok so the node is on with this config

–socket-path /opt/cardano/cnode/sockets/node0.socket

But in the folder /opt/cardano/cnode/sockets/ has zero contents

What is happening.

Is it quicker I just trash this VM and build a node from source ?

Its crazy as I followed all CNtools instructions the node synced all blocks for 12 hours then stopped. ?

Should I report this to anywhere else ?

Is this just on certain servers ?

Hi what are the best doc links to follow to build Stake pools from source ?

Hi Think it is all here

Hi as suggested please run free -m 8GB RAM is usually not enough. If that is the problem you will add a swap partition to try to extend the RAM that way. Official docs say the minimum requirement is 12GB for 1.33.0. 16GB would be nice to have.

If you are using UFW please run sudo ufw status numbered to see if you allow traffic for the configured port.

Check first if RAM is the issue.

This one: https://www.coincashew.com/coins/overview-ada/guide-how-to-build-a-haskell-stakepool-node

Yes I am now rebuilding new bigger VMs following that guide on Section (4)

I am just doing Typology.json file Does this look right with 2 relays for the Producer ?

cat > $NODE_HOME/${NODE_CONFIG}-topology.json << EOF

{

“Producers”: [

{

“addr”: “x.x.x.x”,

“port”: 6000,

“valency”: 1

}

{

“addr”: “x.x.x.x”,

“port”: 6000,

“valency”: 1

}

]

}

EOF

Would you recommend changing the port 6000 later ?

Looks ok, but you are missing a comma so the json is not valid.
The website here will tell you the exact spot where the comma is missing: https://jsonformatter.curiousconcept.com/#

No need to change the port, there is no real gain. Just make sure to configure the firewall. It’s all very well documented in the CoinCashew guide. There is a separate section “Hardening your ubuntu server” which takes you through setting up the firewall and other security related settings.

Hi I am using this port checker and it says 6000 is closed on the relay I have a GCP Firewall rule setup with 6000/tcp and on the instance its on UFW anywhere else I need it on ?

udo ufw status
Status: active

To Action From


22/tcp ALLOW Anywhere
6000/tcp ALLOW Anywhere
22/tcp (v6) ALLOW Anywhere (v6)
6000/tcp (v6) ALLOW Anywhere (v6)

Thanks I added the comma

This means the 6000 port is open for all IPs. It will work but from the security aspect it’s not good.

Ok, but set to “LIMIT” to block IPs which are connecting often within a small time interval (like brute force hacking the password)

You should allow connections just from the relays. No other IP should be able to connect.

As UFW is not blocking, seems GCP is misconfigured.

Yes of course I was allowing all only for testing ports

Yes must be GCP I think too.