Single Node Deployment
This document shows you how to install LogScale on a single Ubuntu Server 18.04 or later LTS node. Production environments should be clustered for failover and redundancy purposes.
For a simpler single node deployment that uses Docker, see Docker Deployment.
For multi-server installation, please see Cluster Management.
In order to install and run LogScale on a single server, the server will requires a minimum of 16 CPU cores, 16 GB of memory, and a 1 GBit Network card. Disk space will depend on the amount of ingested data per day and the number of retention days. This is calculated as follows:
Retention Days x GB Injected / Compression Factor
That will determine the needed disk space for a single server.
On Amazon's AWS, for a single server, start with Ubuntu M5.4XL. This instance type contains 16 vCPUs, 64 GB memory, up to 10 Gbps network). Plus, using AWS with Ubuntu is the easiest method for installing LogScale.
In addition to port 22 for SSH, the LogScale node requires port 8080 opened to incoming traffic for handling requests by the web application (i.e., the LogScale User Interface) and API. If the node is to be part of a cluster, it will need to have TCP incoming ports 8080 and 9200 open, although 9200 is optional.
As for software in addition to whatever is used by the operating system and normal server functions, you'll have to install the following:
LogScale runs on the Java Virtual Machine. You'll have to install JVM 11.2+3 before installing LogScale.
LogScale uses Apache Kafka internally for queuing incoming messages, and for storing shared state when running LogScale in a cluster setup. You'll need to install Apache Kafka 2.4+.
The last software you'll need to install is the LogScale software. Click on the heading here to see preparation, download, and installation instructions.