I am currently testing temBoard on a 3-node PostgreSQL cluster managed by Patroni. After installing the agent on the first node, the UI treats it as a standalone instance. Furthermore, when I installed the agent on the second node, the service repeatedly reported errors until I manually disabled the replication_lag probe in the configuration.
This was an error:
root@IEH-P2CLUSTER2:~# journalctl -u temboard-agent@ieh_cluster_2.service -f Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: monitoring: connection to server at "192.168.11.171", port 5432 failed: fe_sendauth: no password supplied Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: Traceback (most recent call last): Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/temboardagent/plugins/monitoring/probes.py", line 739, in run Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: psycopg2.connect(dsn, connection_factory=PhysicalReplicationConnection) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/psycopg2/__init__.py", line 122, in connect Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: conn = _connect(dsn, connection_factory=connection_factory, **kwasync) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/psycopg2/extras.py", line 507, in __init__ Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: super().__init__(*args, **kwargs) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: psycopg2.OperationalError: connection to server at "192.168.11.171", port 5432 failed: fe_sendauth: no password supplied Mar 26 14:15:48 IEH-P2CLUSTER2 temboard-agent[3838508]: dashboard: Collecting metrics. Mar 26 14:15:50 IEH-P2CLUSTER2 temboard-agent[3838508]: dashboard: Collecting metrics.
Is temBoard designed to handle high-availability cluster architectures like Patroni, or is it primarily intended for monitoring independent standalone instances?
Thank you for your help!
I am currently testing temBoard on a 3-node PostgreSQL cluster managed by Patroni. After installing the agent on the first node, the UI treats it as a standalone instance. Furthermore, when I installed the agent on the second node, the service repeatedly reported errors until I manually disabled the replication_lag probe in the configuration.
This was an error:
root@IEH-P2CLUSTER2:~# journalctl -u temboard-agent@ieh_cluster_2.service -f Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: monitoring: connection to server at "192.168.11.171", port 5432 failed: fe_sendauth: no password supplied Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: Traceback (most recent call last): Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/temboardagent/plugins/monitoring/probes.py", line 739, in run Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: psycopg2.connect(dsn, connection_factory=PhysicalReplicationConnection) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/psycopg2/__init__.py", line 122, in connect Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: conn = _connect(dsn, connection_factory=connection_factory, **kwasync) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: File "/usr/lib/python3/dist-packages/psycopg2/extras.py", line 507, in __init__ Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: super().__init__(*args, **kwargs) Mar 26 14:15:46 IEH-P2CLUSTER2 temboard-agent[3838509]: psycopg2.OperationalError: connection to server at "192.168.11.171", port 5432 failed: fe_sendauth: no password supplied Mar 26 14:15:48 IEH-P2CLUSTER2 temboard-agent[3838508]: dashboard: Collecting metrics. Mar 26 14:15:50 IEH-P2CLUSTER2 temboard-agent[3838508]: dashboard: Collecting metrics.Is temBoard designed to handle high-availability cluster architectures like Patroni, or is it primarily intended for monitoring independent standalone instances?
Thank you for your help!