Amount of lookup requests sent into the desired destination domain. Use to check if the load of cross-cluster search requests are frustrating your domain, correlate any spike During this metric with any JVM/CPU spike.
For a particular link, the sum of follower checkpoint values throughout all replicating indexes. You should utilize this metric to measure replication latency.
The number of situations that "youthful generation" garbage selection has operate on UltraWarm nodes. A significant, ever-escalating quantity of runs is a traditional Element of cluster functions.
For those who deploy your facts instances in just one AZ, your committed learn occasions can also be deployed in exactly the same AZ. However, when you deploy your details occasions across two or three AZs, Amazon OpenSearch Assistance mechanically distributes the devoted grasp circumstances throughout three AZs.
which include time used inside the queue. This value could be the sum in the amount of time it's going to take to finish the power merge, snapshot, and shard relocation levels with the migration approach.
Cluster configuration variations might interrupt these functions before completion. We propose that you make use of the /_tasks Procedure together with these operations to confirm which the requests done effectively.
Amazon OpenSearch Services publishes knowledge from your domains to Amazon CloudWatch. CloudWatch lets you retrieve data about People knowledge factors as an ordered established of time-series info, referred to as metrics
A value of one indicates that the primary shards for all indexes are allocated to nodes within the cluster, but duplicate shards for a minimum of a single index are not. To find out more, see Yellow cluster status.
The number of turned down jobs in the bulk thread pool. If this quantity continually grows, take into account scaling your cluster.
Replicate indexes, mappings, and metadata from just one OpenSearch cluster to another so that you can develop cross cluster redundancy or offload reporting querying to the secondary cluster.
The indexed info are JSON paperwork containing precise fields including state (the region from which an software is invoked), message (which will comprise valuable information and facts for instance a lambda payload), correlationID
A worth of one implies OpenSearch support that swapping due to page faults has probably caused spikes in underlying disk use through a particular time frame. Suitable stats: Utmost
The utmost share of your Java heap employed for all information nodes while in the cluster. OpenSearch Services employs half of the occasion's RAM to the Java heap, as many as a heap dimension of 32 GiB.
The amount of time, in milliseconds, which the cluster has used undertaking "previous technology" rubbish selection.