cardano-node
cardano-node copied to clipboard
[BUG] - Submit api "Network.Socket.bind: resource busy (Address in use)"
External
Area Other Any other topic (Delegation, Ranking, ...).
Summary
When trying to run two instances(preprod+preview) of cardano-submit-api with the same IP but different ports, one of them throws the error: Network.Socket.bind: resource busy (Address in use).
Steps to reproduce
- Start cardano-node in the preview network on 127.0.0.1:5003
- Start cardano-node in the preprod network on 127.0.0.1:5002
- Start submit-api in the preview network on 127.0.0.1:8997
- Start submit-api in the preprod network on 127.0.0.1:8998
- Check the console.
Expected behavior The check for availability should take the port into account. Both instances should be able to run in parallel.
System info (please complete the following information):
- OS Name: Ubuntu
- OS Version: 22.04
- Node version
cardano-node 9.1.1 - linux-x86_64 - ghc-8.10
Additional context Two cardano-node instances for different networks are running on 127.0.0.1 on different ports. The cardano-submit-api instances are also running on 127.0.0.1 on different ports. No matter the order in which they are started, only the one that was started first will work. I don't remember this behavior occurring in previous versions.
Solution: the --metrics-port argument should be specified with a unique number.
But I still think that it shouldn't be a stopper. Maybe there should be a warning, like the Node does, but not a crash.
We are able to reproduce the issue as reported when running multiple node instances but using different port numbers.
Proper error messages or documentation should be provided to include metrics-port, or even better; we should be able to run in parallel with just enough configuration.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
pewpew
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
.
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 120 days.
bump
@carbolymer - would you know if this is even on the radar? shouldnt be a huge task, feels a bit weird to keep bumping for close to a year without any reaction
We haven't forgot about it. I'll have a look this week.