The InfluxDB schema updater is a small DevOps tool to manage the schema of an InfluxDB instance with a set of configuration files.

influxdb-schema-updater [--help] [--dryrun] [--diff] [--force] [--config <schema_location>] [--url <url>]

• --help

  Print a help message and exit.

• --dryrun

  Print the changes which would be applied in normal mode.

• --diff

  Print the InfluxQL queries instead of executing them.

• --force

  Apply the changes which were prevented in normal mode.

• --config

  The directory where the schema files are located. Default is /etc/influxdb/schema/.

• --url

  The url where the InfluxDB HTTP API is reachable. Default is localhost:8086.

This tool compares the databases, retention policies (RPs) and continuous queries (CQs) found in the <schema_location> directory to the ones in the InfluxDB instance reachable at <url>. If there is a difference, InfluxDB will be updated. Some changes like deleting a database are skipped when the --force flag is not set.

The exit code is 0 if and only if every required update has been executed successfully.

The <schema_location> directory should have the following structure:

db/
    <db_file1>
    <db_file2>
    ...
cq/
    <cq_file1>
    <cq_file2>
        ...

The files in db/ contain CREATE queries for databases followed by their their RPs, for example:

CREATE DATABASE test WITH DURATION 100d REPLICATION 1 SHARD DURATION 2w NAME rp1;
CREATE RETENTION POLICY rp2 ON test DURATION 260w REPLICATION 1 SHARD DURATION 12w;

CREATE DATABASE test2;
CREATE RETENTION POLICY rp1 ON test2 DURATION 100d REPLICATION 1 SHARD DURATION 2w;
CREATE RETENTION POLICY rp2 ON test2 DURATION 260w REPLICATION 1 SHARD DURATION 12w;
CREATE RETENTION POLICY rp3 ON test2 DURATION INF REPLICATION 1 SHARD DURATION 260w;

The files in cq/ contain CQs usually corresponding to the databases declared in the file with the same name in db/, for example:

CREATE CONTINUOUS QUERY cq1 ON test RESAMPLE EVERY 5m FOR 10m BEGIN SELECT LAST(a) AS b, c INTO test.rp2.m FROM test.rp1.m GROUP BY time(5m) END;

CREATE CONTINUOUS QUERY cq1 ON test2 RESAMPLE EVERY 30m FOR 1h BEGIN SELECT LAST(a) AS b, c INTO test2.rp2.m FROM test2.rp1.m GROUP BY time(30m) END;
CREATE CONTINUOUS QUERY cq2 ON test2 RESAMPLE EVERY 1d FOR 2d BEGIN SELECT LAST(a) AS b, c INTO test2.rp3.m FROM test2.rp2.m GROUP BY time(1d) END;