iptables: No chain/target/match by that name.

[ni3gavhane]

Issue:
Some validator organization are commenting following lines in the installation scripts which results into unexpected end of the installation script.
line 65 to 68

echo "Securing SSH..."
writeSShConfig
semanage port -a -t ssh_port_t -p tcp 2222
service sshd restart

Result:
parity-keygen
parity-keygen
Creating docker-stack_parity-telemetry_1 ... done
Creating docker-stack_nodecontrol_1 ... done
Creating docker-stack_parity_1 ... done
Waiting 30 sec for parity to come up and generate the enode...
Redirecting to /bin/systemctl restart telegraf.service
Setting up firewall
iptables: No chain/target/match by that name.
iptables: Directory not empty.
iptables: Chain already exists.
iptables: No chain/target/match by that name.

Expected Outcome:
Script should end with a proper installation summary.

Details:
Installation Script:
CentOS: https://github.com/energywebfoundation/ewc-validator-node-install-scripts/blob/master/volta-affiliate/install-validator-centos-7-volta.sh
Ubuntu: https://github.com/energywebfoundation/ewc-validator-node-install-scripts/blob/master/volta-affiliate/install-validator-ubuntu-server-18.04-volta.sh

[kazmen1983]

which distribution you are using? Is this CentOS ?

[ni3gavhane]

One of the member has this problem on CentOS.

[kazmen1983]

it looks like iptables problem. Can they start script with the bash -x ./install-validator-centos-7-volta.sh and copy output?

[Juanyyo82]

Hi,
Is a virtual server in a local ESX server, with ubuntu 18.04.2 (install-validator-ubuntu-server-18.04-volta.sh). I did the instalation installation from scratch, fresh ubuntu server and full update OS.
I have tu put in line 178 "--no-check-certificate", Because we have a proxy server and the installation display this error: "ERROR: cannot verify raw.githubusercontent.com's certificate, issued by ‘CN=Myproxy’:"
Is there any problem to install the centos script in a rhel 7 OS?
Regards

[bartcoelus]

Hi. It was me facing this issues. First of all important: we configured our SSH on port 2222. The script reconfigures SSH - but can't handle if the config is already set to 2222.
Next to that we had a hard time with IPtables

iptables: No chain/target/match by that name.
iptables: Directory not empty.
iptables: Chain already exists.
iptables: No chain/target/match by that name.

[bartcoelus]

still unsolved... waiting for further instructions - Rodger

[kazmen1983]

I was asking for the running script with bash -x so I can look which Iptables rules are failing.

[bartcoelus]

docker: Error response from daemon: driver failed programming external connectivity on endpoint parity-keygen (59f......): Bind for 127.0.0.1:8545 failed: port is already allocated.

[bartcoelus]

not sure if this is due a previous version/deamon running... also I removed the key-info from this post (59f....) since I don't want to reveal anything harmfull in public

[ni3gavhane]

@bartcoelus can you please post the complete log. you can do something like following:
./installation_script.sh 2>&1 | tee debug.log

and share the debug.log

Also, it seems the parity instance is already running that is why you get the last error.
check that with the following command:

ps aux | grep parity

if you see processes running, kill them with kill -9 <PID>

then remove the docker-stack folder rm -rf docker-stack and re-run the script.

[bartcoelus]

done that - can I send you the debug.log somewhere outside github. I don't think we want all this stuff on this public page

[kazmen1983]

Client sended e-mail that now validator install works.

[bartcoelus]

Let's share important learnings from our case:

1. The script defines port 2222 - since our firewall and SSH was already configured to use that port we needed to comment these lines. Maybe a good idea to throw-catch the event if the connection is already on that port.
2. Needed to kill the docker containers in memory in order to clean up a failed install
3. Important is to make sure the user operates from a root directory and not to install as default root
4. We'd not recommend to name the node with IP-adress as default suggested in the script. When the IP adress ever changes it can cause confusion.
5. Removal of docker-directories and spool files is needed in the case you need to redo the install. I refer to the wiki - it contains the directories and files that need to be removed at a failed attempt.
   https://energyweb.atlassian.net/wiki/spaces/EWF/pages/702513299/Volta+Setting+Up+a+New+Validator+Node