Configuration Reference
Use the following properties, found in ${OPENNMS_HOME}/etc/opennms.properties
, to configure Newts.
General
Name | Description | Default |
---|---|---|
org.opennms.newts.config.keyspace |
Name of the keyspace to use. |
newts |
org.opennms.newts.config.hostname |
IP address or hostnames of the Cassandra nodes. If including multiple node references, separate each with a single comma. |
localhost |
org.opennms.newts.config.port |
CQL port used to connect to the Cassandra nodes. |
9042 |
org.opennms.newts.config.username |
Username to use when connecting to Cassandra via CQL. |
cassandra |
org.opennms.newts.config.password |
Password to use when connecting to Cassandra via CQL. |
cassandra |
org.opennms.newts.config.datacenter |
Set this to the value of your local datacenter for the best performance. |
datacenter1 |
org.opennms.newts.config.ssl |
Enable or disable SSL when connecting to Cassandra. |
false |
org.opennms.newts.config.core-connections-per-host |
Core number of connections per host. |
Driver default |
org.opennms.newts.config.max-connections-per-host |
Maximum number of connections per host. |
Driver default |
org.opennms.newts.config.max-requests-per-connection |
Maximum number of requests that can be in flight on a single connection at the same time. |
Driver default |
org.opennms.newts.config.read_consistency |
Consistency level for read operations. See Configuring Data Consistency in the Cassandra documentation for a list of available options. |
ONE |
org.opennms.newts.config.write_consistency |
Consistency level for write operations. See Configuring Data Consistency in the Cassandra documentation for a list of available options. |
ANY |
org.opennms.newts.config.max_batch_size |
Maximum number of records to insert in a single transaction.
Limited by the size of the Cassandra cluster’s |
16 |
org.opennms.newts.config.ring_buffer_size |
Maximum number of records that can be held in the ring buffer. Must be a power of two. |
8192 |
org.opennms.newts.config.writer_threads |
Number of threads used to pull samples from the ring buffer and inserted into Newts. |
16 |
org.opennms.newts.config.ttl |
Number of seconds after which samples will automatically be deleted. Defaults to one year, in seconds. |
31540000 |
org.opennms.newts.config.resource_shard |
Time (in seconds) that samples are stored at the same key. Defaults to 7 days, in seconds. |
604800 |
org.opennms.newts.query.minimum_step |
Minimum step size, in milliseconds. Used to prevent large queries. |
300000 |
org.opennms.newts.query.interval_divider |
If the query does not specify an interval, the step will be divided into this number of intervals when aggregating values. |
2 |
org.opennms.newts.query.heartbeat |
Duration, in milliseconds. Used when no heartbeat is specified. Should generally be 1.5 times your largest collection interval. |
450000 |
org.opennms.newts.query.parallelism |
Maximum number of threads that can be used to compute aggregates. Defaults to the number of available cores. |
Number of cores |
org.opennms.newts.config.cache.strategy |
Canonical name of the class used for resource-level caching. See the table below for all of the available options. |
See below |
org.opennms.newts.config.cache.max_entries |
Maximum number of records to keep in the cache when using an in-memory caching strategy. |
8192 |
org.opennms.newts.nan_on_counter_wrap |
Disables the processing of counter wraps, replacing them with NANs instead. |
false |
org.opennms.newts.config.cache.priming.disable |
Disables the cache primer, which preemptively loads the cache with indexed resources on startup. |
false |
org.opennms.newts.config.cache.priming.block_ms |
Block startup for this many milliseconds while waiting for the cache to be primed.
Set this value to |
120000 |
Name | Class | Default |
---|---|---|
In-memory cache |
org.opennms.netmgt.newts.support.GuavaSearchableResourceMetadataCache |
Y |
Redis-based cache |
org.opennms.netmgt.newts.support.RedisResourceMetadataCache |
N |
Redis Cache
When enabled, the following options can be used to configure the Redis-based cache:
Name | Description | Default |
---|---|---|
org.opennms.newts.config.cache.redis_hostname |
IP address the Redis server’s hostname. |
localhost |
org.opennms.newts.config.cache.redis_port |
TCP port used to connect to the Redis server. |
6379 |
Recommendations
You will likely want to change the values of cache.max_entries
and ring_buffer_size
to suit your installation.
Meridian caches metadata related to resources to avoid writing redundant records in Cassandra.
If you are collecting data from a large number of resources, you should increase cache.max_entries
to reflect the number of resources you are collecting from, along with a suitable buffer.
The samples the collectors gather are temporarily stored in a ring buffer before they are persisted to Cassandra using Newts.
You should increase the value of ring_buffer_size
if you expect large peaks of collectors returning at once or latency in persisting these to Cassandra.
Note, however, that the memory the ring buffer uses is reserved, and larger values may require an increased heap size.
Newts uses cache priming to help reduce the number of records that need to be indexed after restarting Meridian. This works by rebuilding the cache using the index data that has already been persisted in Cassandra. If you continue to see large spikes of index-related inserts after rebooting, consider increasing the amount of time spent priming the cache.