Backup

The backup script is installed at /opt/energysoar/lib/extra/backup.sh. Run it as root. Output files are written to /root/.

Backup modes

The script supports three modes:

Flag	Mode	What is backed up
(none)	Config only	Configuration files. Output: `backup_energysoar_config_TIMESTAMP.zip`
`-d`	Data only	Elasticsearch indices, Cassandra data, n8n data. Output: `backup_energysoar_data_TIMESTAMP.zip`
`-a`	Full backup	Config and data combined.

Config backup

Backs up all configuration directories:

/etc/cortex/
/opt/cortex/Energy-Analyzers/
/etc/energysoar/
/var/lib/n8n/.n8n/
/opt/energysoar/n8n/scripts/
/etc/nginx/conf.d/energysoar.conf
/etc/cassandra/

/opt/energysoar/lib/extra/backup.sh

The resulting ZIP is saved as /root/backup_energysoar_config_TIMESTAMP.zip.

Data backup

Exports:

Elasticsearch indices (using elasticdump)
Cassandra keyspace tables (using cqlsh COPY) and schema
n8n workflow data from /var/lib/n8n/.n8n/

/opt/energysoar/lib/extra/backup.sh -d

The resulting ZIP is saved as /root/backup_energysoar_data_TIMESTAMP.zip.

Full backup

Runs config backup followed by data backup:

/opt/energysoar/lib/extra/backup.sh -a

HA mode

In a high-availability cluster, pass the -H flag. The script will prompt for Cassandra credentials before exporting cluster data:

/opt/energysoar/lib/extra/backup.sh -H
/opt/energysoar/lib/extra/backup.sh -a -H

Credentials can also be passed via environment variables to avoid interactive prompts:

CASS_USER=energysoar CASS_PASS=password /opt/energysoar/lib/extra/backup.sh -a -H

Restore

The restore procedure varies by what failed. Always stop all services before restoring data and restart them afterwards.

Stop services before restore:

systemctl stop energysoar cortex n8n nginx

Restore configuration

Extract the config backup archive to the root of the filesystem:

unzip /root/backup_energysoar_config_TIMESTAMP.zip -d /

This overwrites /etc/cortex/, /etc/energysoar/, /etc/nginx/, /etc/cassandra/, and the n8n scripts directory with the backed-up versions. Verify file ownership after extraction — files must be readable by the energysoar and cortex service users.

Restore data

Extract the data backup archive:

unzip /root/backup_energysoar_data_TIMESTAMP.zip -d /tmp/energysoar-restore

Cassandra — ensure Cassandra is running before importing:

systemctl start cassandra
cd /tmp/energysoar-restore/cassandra/
cqlsh -e "SOURCE 'schema.cql';"
cqlsh -e "COPY energysoar.<table_name> FROM '<table_name>.csv' WITH DELIMITER='|' AND HEADER=TRUE;"

Repeat the COPY command for each .csv file in the directory.

Elasticsearch / Energy Logserver — restore each index using elasticdump:

cd /tmp/energysoar-restore/elasticsearch/
for f in *_mapping.json; do
    index="${f%%_mapping.json}"
    elasticdump --input="${index}_mapping.json" \
                --output="http://127.0.0.1:9200/${index}" \
                --type=mapping
    elasticdump --input="${index}.json" \
                --output="http://127.0.0.1:9200/${index}" \
                --type=data
done

n8n workflows — copy the n8n data directory back:

cp -r /tmp/energysoar-restore/n8n/.n8n/ /var/lib/n8n/
chown -R n8n:n8n /var/lib/n8n/.n8n/

Start services after restore:

systemctl start cassandra elasticsearch energysoar cortex n8n nginx

Verify that all services are healthy using:

systemctl status energysoar cortex n8n

See Diagnostics for health check endpoints to confirm data availability.