The steps to follow in order to restore data on the faulty node:
1)Stop the Hub on all nodes:
- If systemd is managing the Hub:
sudo systemctl mask hub-orchestration
sudo systemctl stop hub-orchestration
- or run `/opt/hub/current/bin/hub.sh stop` on each node
2)Ensure that the Hub is stopped on each node. First run `/opt/hub/latest/bin/hub.sh status` then check that Postgres is effectively stopped: `ps -edf|grep postgres:`
3)On one healthy node, make a tarball of /opt/hub/postgres and a tarball of /opt/hub/pgwal and copy across to node <node>
4)On the faulty node, rename /opt/hub/postgres to /opt/hub/postgres.<timestamp> and /opt/hub/pgwal to /opt/hub/pgwal.<timestamp>.
5)On the faulty node, unzip the tarballs copied from healty node so that you end up with /opt/hub/postgres and /opt/hub/pgwal
6)On the faulty node, start Postgres manually by running `/opt/hub/latest/services/postgres-timescale-12.6-1.7.4/start.sh` and wait until you see the following in the logs:
LOG: database system is ready to accept connections
LOG: TimescaleDB background worker launcher connected to shared catalogs
7)On the faulty node, stop Postgres using a Ctrl-C
8)Restart the Hub normally using `hubctl` or systemd.
If you have any further questions:
- Please contact with our Client Services team via the chat service box available in any of our websites or via email to email@example.com
Please sign in to leave a comment.