The mentioned error is not a problem, but your node should have gotten some traffic.
It’s likely that your node has been paused for some reason. There is now an API you can use to check your reputation.
If the score for audits is below 0.6 your node has been paused on that satellite. You can contact support@storj.io with an explanation of your situation, your node ID and your run command to ask for help if that is the case. But before you do, try to figure out why your node may have been paused. Have you lost data? Have you been offline for long periods of time? If you do contact support include anything that may be relevant.
I have added docker port mapping command and run “curl 127.0.0.1:14002/api/dashboard”, and ended up with “curl: (52) Empty reply from server”.
Below is command insider docker exec to prove port 14002 is enabled,
/app # ps aux
PID USER TIME COMMAND
1 root 3:55 ./storagenode run --config-dir config --identity-dir identity --metrics.app-suffix=-alpha --metrics.interval=30m --kademlia.external-address=61.172.47.198:4511 --kademlia.operator.email=larry.*****@******.com --kademlia.operator.wallet=0x537eB99Acda7E8************* --console.address=:14002 --storage.allocated-bandwidth=20TB --storage.allocated-disk-space=6TB
25 root 0:00 /bin/sh
35 root 0:00 ps aux
Any idea, I am using arm docker image.
Best Regards
Larry Zheng
Just for your info that your ARM docker release seems missing dashboard API info.
cc@cc_0394:~$ curl 127.0.0.1:14002/api/satellite/118UWpMCHzs6CvSgWd9BfFVjw5K9pZbJjkfZJexMtSkmKxvvAW
curl: (52) Empty reply from server
cc@cc_0394:~$
cc@cc_0394:~$ curl 127.0.0.1:14002/api/dashboard
curl: (52) Empty reply from server
I did apple to apple comparison by setting up another AMD64 build, which output correct node info via dashboard API.
You shouldn’t need to expose that port since the wget command is running within the container itself.
Alternatively, if you don’t expose the port, you should be able to hit the API from the docker host using the internal docker IP (172.17.0.something unless you’ve tweaked that).
FYI; This is weird; I’ve just set up two storage nodes; running on almost exactly the same software (Ubuntu 18.04.3 + docker-ce 19.03.1 - pulled same image, has same hash, with -p 127.0.0.1:14002:14002, made identity same way (except for auth token ofcourse) - and on one, my script (same on both) works and can poll api/satellite outside docker, and on the other, it fails, until I comment out the # console.address: 127.0.0.1:14002
line - or else it throws a HTP500 internal server error from outside the container (not within).
Absolutely weird, as I’ve done pretty much exactly the same on both nodes…
Not weird. That line limits it to local Host inside the container. You shouldn’t add this line to the config.yaml. it defaults to :14002 in the current docker image which will work just fine.
You are absolutely correct; what I was wondering was why the same config yields different behavior on identical setups. One works, the other does not - until the config was changed. Same platform, same everything (to the best of my knowledge).
If I remember correctly that setting is passed through the entrypoint in the docker container. I believe that SHOULD overwrite the config.yaml settings. But not entirely sure. Make sure you have the latest :beta tag version of the image on both nodes.