You can deploy vRealize Log Insight with a single node, single cluster, or cluster with forwarders.
External load balancers are not supported for use with vRealize Log Insight, including vRealize Log Insight clusters.
A basic vRealize Log Insight configuration includes a single node. The log sources are applications, OS logs, virtual machine logs, hosts, the vCenter Server, virtual or physical switches and routers, storage hardware, and so on. Log streams are transported to the vRealize Log Insight node using syslog (UDP, TCP, TCP+SSL) or CFAPI (the vRealize Log Insight native ingestion protocol over HTTP or HTTPS), either directly by an application, syslog concentrator, or the vRealize Log Insight agent installed on the source.
It is a best practice for single-node deployments to use the vRealize Log Insight integrated load balancer (ILB) and to send queries and ingestion traffic to the ILB. This does not incur overhead and simplifies configuration if you want to add nodes to create a cluster for your deployment in the future.
Single nodes are not recommended for production environments.
The use of clusters is recommended for production environments. Two-node clusters are not supported. Nodes in clusters should all be of the same size and in the same datacenter. The ILB used with clusters requires that nodes be in the same L2 network.
A minimum of medium-sized nodes are recommended for production environments. If you anticipate working with a high number of concurrent queries, including alerts, consider using large-sized nodes. For information about sizing, see Sizing the vRealize Log Insight Virtual Appliance.
A vRealize Log Insight single cluster configuration includes a minimum of three nodes to twelve nodes and uses the ILB.
A single log message is present in only one location within the cluster. The cluster remains up and available to ingest data and serve queries during the temporary unavailability of any single node results in the cluster. Removal and reintroduction of a cluster node is not supported. A minimum of three healthy nodes is required, or, four nodes to support a single-node failure scenario.
A vRealize Log Insight cluster's virtual IP uses a Linux Virtual Server in Direct Server Return Mode (LVS-DR) for load balancing. Direct Server Return is more efficient than routing all response traffic through a single cluster member. However, it also resembles spoofed traffic, which NSX Distributed Firewall blocks. The vRealize Log Insight virtual machines must be excluded from VMware NSX Distributed Firewall Protection.
Cluster with Forwarders
A vRealize Log Insight cluster with forwarders configuration includes main indexing, storage, and a query cluster of three to twelve nodes using the ILB. A single log message is present in only one location within the main cluster, as for the single cluster.
The design is extended through the addition of multiple forwarder clusters at remote sites or clusters. Each forwarder cluster is configured to forward all its log messages to the main cluster and users connect to the main cluster, taking advantage of CFAPI for compression and resilience on the forwarding path. Forwarder clusters configured as top-of-rack may be configured with a larger local retention.
Cross-forwarding Central for Redundancy
This vRealize Log Insight deployment scenario includes a cluster with forwarder that is extended and mirrored. Two main clusters are used for indexing, storage, and query. One main cluster is in each data center. Each is front-ended with a pair of dedicated forwarder clusters. All log sources from all top-of-rack aggregations concentrate at the forwarder clusters. You can independently query the same logs on both retention clusters.
vRealize Log Insight Integrated Load Balancer (ILB)
To properly balance traffic across nodes in a cluster and to minimize administrative overhead, use the ILB for all deployments. This ensures that incoming ingestion traffic is accepted even if some vRealize Log Insight nodes are unavailable.