Environment

• YugabyteDB Anywhere - All versions

Issue

• YugabyteDB Anywhere is unavailable on a Replicated-based installation or metrics within Platform are unavailable.

Resolution

Overview

To troubleshoot YugabyteDB Anywhere when it has become unavailable, first check if a log file has grown too large, check if the Prometheus metrics are taking up too much disk space, restart the Replicated containers, and gather logs for a Support ticket, if necessary.

Steps

Check if log file has grown too large

1. As root on the YugabyteDB Anywhere (platform)  node:

# cd /var/lib/docker/containers

2. Check if there is a the directory that is several GB in size (example output below)

# du -sh *

3. Change into the large directory (if exists)

# cd <large_directory>

4. Delete the log file that is several GB in size

# rm <container_long_id>-json.log

5. Restart the docker daemon

# systemctl restart docker

Check if Prometheus metrics are filling up the disk space

1. See the metrics KB: How to troubleshoot high disk utilization issues for replicated based yugaware install

Restart Replicated Containers

1. Check container status

sudo docker ps

2 Restart Replicated containers (Replicated, Operator, UI)

sudo docker restart replicated

sudo docker restart replicated-operator

sudo docker restart replicated-ui

Gather Logs for Yugabyte Support Ticket

1. Get the YugabyteDB Anywhere logs (Note: a container restart will erase these)

sudo docker logs yugaware

2. Get the Postgres logs

sudo docker logs postgres

3. Get the Prometheus logs:

sudo docker logs prometheus