AliES is a highly tailored kernel for Alibaba Cloud Elasticsearch. AliES supports all the features provided by the open source Elasticsearch kernel and provides additional features, such as metric optimization, thread pooling, circuit breaking optimization, and query and write performance optimization. These additional features are developed based on the abundant experience of the Alibaba Cloud Elasticsearch team in multiple scenarios. These features not only improve cluster stability and performance but also reduce costs and extend the scope of monitoring and O&M. This topic describes the optimized features in each version of AliES.

Elasticsearch V7.10.0 of the Standard Edition

Kernel version Description
V1.4.0
  • The aliyun-knn plug-in is updated to improve write performance. The plug-in supports script queries and is integrated with the optimized capabilities of the related hardware to improve the vector search feature.
  • The aliyun-qos plug-in is optimized to improve cluster-level throttling. When you use this plug-in, you do not need to focus on the topology and load of the nodes in your Elasticsearch cluster. Traffic is automatically distributed to the nodes. This improves cluster usability and stability.
V1.3.0
  • The slow query isolation feature is provided to reduce the impact of anomalous queries on cluster stability.
  • The gig plug-in is provided to perform a switchover within seconds after an exception occurs on a cluster. This plug-in prevents query jitters caused by anomalous nodes.
    Note For Elasticsearch V7.10.0 clusters of the Standard Edition, the gig plug-in is integrated into the aliyun-qos plug-in. The aliyun-qos plug-in is installed by default.
  • The physical replication feature is provided to improve the write performance of indexes that have replica shards.
  • The pruning feature is provided for time series indexes to improve the query performance of the indexes.
  • The access logs of clusters can be viewed. These logs contain fields such as Time, Node IP, and Content. You can use these logs to troubleshoot issues and analyze requests.
  • The scheduling performance of dedicated master nodes is improved by 10 times. Each dedicated master node can schedule more shards.

Elasticsearch V6.7.0 of the Standard Edition

Kernel version Description
V1.3.0
  • The slow query isolation feature is provided to reduce the impact of anomalous queries on cluster stability.
  • The gig plug-in is provided to perform a switchover within seconds after an exception occurs on a cluster. This plug-in prevents query jitters caused by anomalous nodes.
Notice Before you use the preceding features, make sure that the kernel version of your Elasticsearch cluster is V1.3.0. Otherwise, upgrade the kernel. You can upgrade only the kernels of Standard Edition clusters whose kernel versions are V0.3.0, V1.0.2, or V1.3.0.
V1.2.0
  • The physical replication feature is provided to improve the write performance of indexes that have replica shards.
  • The pruning feature is provided for time series indexes to improve the query performance of the indexes.
  • Primary key-based data deduplication is optimized during queries. This improves the write performance of documents that contain primary keys by 10%.
  • Finite state transducers (FSTs) that do not occupy heap memory are supported. A single node can store a maximum of 20 TiB of index data.
V1.0.2 The access logs of clusters can be viewed. These logs contain fields such as Time, Node IP, and Content. You can use these logs to troubleshoot issues and analyze requests.
V1.0.1 Circuit breaking policies can be configured for Java virtual machines (JVMs). When the usage of the JVM heap memory of your cluster reaches 95%, the system rejects requests to protect the cluster. The following parameters are used to configure the policies:
  • indices.breaker.total.use_real_memory: The default value is false.
  • indices.breaker.total.limit: The default value is 95%.
V0.3.0