Added ServerAliveInterval and ServerAliveCountMax both for node provider's command runner and ssh commands
Added ClientAliveInterval and ClientAliveCountMax in sshd_config to reflect similar keep alive (it is suggested to keep them lower than Server* counterparts).
Enabled integration tests on pull requests (yay!)
Fixed problem, where YagnaService exceptions would be not caught properly at cluster bootstrap
Changed emerging strategy in dev_yaml tool, to be able to remove outbound URLs for goth tests (currently outbound is not supported on goth)
Bumped goth to 0.16.0 to support holesky and yagna in beta versions
Uploaded new image version
Notable highlights:
GCS / SSH commands problems are related to GolemNodeProvider being locked at __init__, while bootstrapping cluster and running yagna fund that in some cases are not working. This will be addressed shortly.
What I've done:
ServerAliveIntervalandServerAliveCountMaxboth for node provider's command runner and ssh commandsClientAliveIntervalandClientAliveCountMaxin sshd_config to reflect similar keep alive (it is suggested to keep them lower thanServer*counterparts).dev_yamltool, to be able to remove outbound URLs for goth tests (currently outbound is not supported on goth)Notable highlights:
__init__, while bootstrapping cluster and runningyagna fundthat in some cases are not working. This will be addressed shortly.