Faster than fast, it’s index-free

Recapping our Quick Start Workshop with James Mountifield

May 25th, 2020

In our Quick Start Workshop, James Mountifield, Humio Principal Solutions Architect (EMEA), presented a demo that showed how to get started using Humio and get the most out of its live streaming search results from the very beginning of your free trial.

We’re saying it’s going to use 20% of current infrastructure, it’s going to run 10x faster, and it’s going to cost less money to buy and operate. The only way we get people to believe it is to have people get their own data in there and see for themselves.”

Watch on-demand

After introducing Humio and explaining how its index-free design contributes to low cost, high-performant machine data aggregation and search, James highlights the most useful resources available for new users of Humio including:

Next, James goes over various installation options for deploying Humio and gives advice based on each.

Humio Installation Options:

  • Bare metal - You can deploy just one node, but he recommends three nodes for production. Follow our guide for Ansible deployments or use Terraform.

  • Cloud Services - Humio works with a variety of cloud providers including AWS EC2 and GCP.

  • Kubernetes - Follow the Helm chart.

  • On Humio Cloud services - Offers a solution without any administrative work.

  • On Docker - Use a Humio-provided Docker image.

He then demonstrates a Docker installation, grabbing a Docker run command from Humio’s docs site and setting up Humio in seconds.

With Humio up and running, James shows viewers a sample dashboard displaying live streaming public data taken from GitHub’s API. He then gives a walkthrough of the central organization page which shows all data repositories, and also drills down into the features on customizations found on each tab, including Search, Dashboards, Alerts, Parsers, and Files.

Exploring Search

James conducts an in-depth query on the data set showing:

  • Humio can search 2.5 million data points in under a second.

  • Even with a complex search that looks over 30 days of data, it scans 1.5 million events per second.

  • Compared to other log management, searches that may take 40 minutes take 40 seconds in Humio.

If you’ve used Splunk or Elastic, you’ll find the query terms very easy to use. The search operators use very standard terms – pipelining language.

Exploring Dashboards and Alerts

From queries, James makes an easy transition to both alerts and widgets on dashboards. Alerts use the same syntax as the search — you can describe thresholds however you want. Simply save a search as a widget and then navigate to the dashboard tab where you can view all your widgets with up-to-the-second accurate data. The speed of Humio’s searches transfers over to increased performance speed for real-time alerts.

We’ve heard people who are limited to 50-100 queries on Splunk. We have some customers running thousands of alerting queries.

Exploring Parsers

James reveals that Humio’s ability to compress data and search at high speed is based on a minimal architecture that features timestamps. Parsers take on a role of added importance in Humio because they extract the timestamp information from all incoming data. If data doesn’t come with timestamp information, Humio applies one automatically based on the arrival time of the event.

Exploring Files

James shows how to enrich incoming data. Instead of wasting processing power on enriching data as it comes in, Humio can add that data at query time. This optimization continues the efficiency gains that Humio has made across all parts of the architecture to support instant ingest and streaming access to data.

Explore the split-second search results for yourself by signing up for a free Humio demo.

To take part in our next Fast Start workshop and have your questions answered live by Humio staff, sign up for our next workshop.

To learn more about how these platforms work together to provide a detailed live streaming picture of application performance, watch the full on-demand video of the last workshop.