orb
orb copied to clipboard
Agents: agent with backend on heartbeat and empty metadata
We're able to see this failure on pipeline:
https://qa.ops.orb.live/orb/staging/d845ba21/orb/python-test/output/#scenario=22042-2557372 https://qa.ops.orb.live/orb/staging/ec357e01/orb/python-test/output/#scenario=41501-18740288
{
"id":"be3fbdd7-5cb8-4dbe-a57e-e6f28e580050",
"name":"test_agent_name_UyVhGuBAkZ",
"state":"online",
"channel_id":"78c3ee7c-350a-4932-b9b4-70b444009c4e",
"agent_tags":{
},
"orb_tags":{
"test_tag_PmuGXH":"test_tag_fJIQ"
},
"agent_metadata":{
},
"last_hb_data":{
"backend_state":{
"pktvisor":{
"state":"running"
}
},
"group_state":{
},
"policy_state":{
}
},
"ts_created":"2022-09-09T18:01:45.761921Z",
"ts_last_hb":"2022-09-09T18:02:12.739794Z"
}
Other agent:
{
"id":"4d670e48-a86a-43c1-a805-02fb483729da",
"name":"test_agent_name_MajLXPkLfy",
"state":"online",
"channel_id":"c846d7d7-9c48-4dcd-bc74-f26a9ad6ce59",
"agent_tags":{
},
"orb_tags":{
"test_tag_xeSDwA":"test_tag_znww"
},
"agent_metadata":{
},
"last_hb_data":{
"backend_state":{
"pktvisor":{
"state":"running"
}
},
"group_state":{
},
"policy_state":{
}
},
"ts_created":"2022-09-12T14:50:32.335276Z",
"ts_last_hb":"2022-09-12T14:51:03.510179Z"
}
The problem is that we have the agent online and the backend properties are correct on agent heartbeat, but the agent metadata is empty. It must NEVER occurs.
In both cases the log [visor] [info] web server listening on localhost:port is displayed twice, and when this failure not happen it appears only once. Maybe not related.. But.. I notice it.