IoTDB uses the multi-replica deployment architecture to implement cluster high availability. Each region (DataRegion and SchemaRegion) has three replicas by default. You can also configure more replicas. If a node is faulty, replicas on other nodes of the region replica can take over services from the faulty node, ensuring service continuity and improving cluster stability.
Parameter |
Role |
Example Value |
Description |
---|---|---|---|
read_consistency_level |
ConfigNode |
strong |
Read consistency level of the custom parameter confignode.customized.configs. Currently, the value can only be strong or weak. |
flush_proportion |
IoTDBServer |
0.4 |
Write memory ratio for invoking disk flushing. If the write load is too high (for example, batch processing = 1000), you can reduce the value. |
replica_affinity_policy |
IoTDBServer |
random |
When the value of read_consistency_level is weak, the strategy of the region replica node is selected for the query task. |
coordinator_read_executor_size |
IoTDBServer |
20 |
Number of read thread cores of the IoTDBServer Coordinator of the custom parameter engine.customized.configs |
rpc_thrift_compression_enable |
ALL |
false |
Whether to compress data during transmission. Data is not compressed by default. |
root.log.level |
ALL |
INFO |
IoTDB log level. The modification of this parameter takes effect without restarting related instances. |
SSL_ENABLE |
ALL |
true |
Whether to encrypt the channel between the client and server using SSL |