Logging and Log Files

Log directories vary by deployment type and by application.

XCO logs are saved to the following locations:

Non-TPVM deployments: /var/log/efa. The installation logs in the /var/log/efa/installer directory are a good source for discovering the reason for a failure.
TPVM deployments: /apps/efa_logs
Kubernetes log files: /var/log/pods
K3s service is available on single and multi-node deployments. The log contains journalctl log of k3s for the past one day.
Keepalived service log files: /var/log/keepalived. The directory contains keepalived service logs. You can use keepalived service logs to debug failovers, double faults, and gateway connectivity. The log contains
- Journalctl log of keepalived for the past one day
- Keepalived service logs
Keepalived service is only available in multi-node deployments. Keepalived has three states: MASTER, BACKUP and FAULT. The active node must have the status of MASTER, the standby will have BACKUP, and FAULT can be seen during a failover.

In multi-node, high availability deployments, logs are replicated on all nodes in the cluster.

The efa system supportsave script gathers all logs, database dumps, pod logs, deployment details, and system support-save and then compresses them into a ZIP folder. You can share this ZIP folder with Extreme support personnel when troubleshooting an issue.

It captures logs of micro-services in XCO, services deployed on the host, and a snapshot of the database. The logs from executing the Supportsave command can be found in the monitoring service logs located at /apps/efa_logs/monitor/ or /var/log/efa/monitor/. Ensure there are no errors during the generation of the Supportsave. If any issues arise during the Supportsave generation, try restarting the monitoring service on the active node.