Humio is a flexible log management solution. We have customers that use our cloud solution, on-prem solutions, and hybrids of both options. Consider what works best for your application and organization and we are happy to help you ind the set up to best suit your needs.
Moving to Humio is easy! We have several common integrations to bring your logs into Humio and we even have a guide on moving from Elastic Stack to Humio - it’s as easy as following a few steps to getting your logs flowing.
Humio was built with containerization in mind! With integrations and existing setup for Kubernetes, Humio is a solution focused on modern deployment solutions.
Humio integrates with several common notification methods including email, Slack, and external services like OpsGenie. If you need Humio to work with your particular notification system, contact our support team!
“Repository” is the new term. What used to be a “dataspace” in Humio is now a Repository.
The HTTP API includes the path
/api/v1/dataspaces/$REPOSITORY_NAME/ to be compatible with existing clients.
In this context, the
$REPOSITORY_NAME variable is the name of the repository. (It used to be the name of the dataspace).
Humio runs on either or both IP versions, depending on what you specify using
HUMIO_JVM_ARGS. By default the process binds on both IPv4 and IPv6.
If you use the Docker images provided by Humio for Kafka and Zookeeper, or run the “humio/humio” image that includes both of them, you need to make sure those processe also get the same options regarding IP protocol as the Humio process.
HUMIO_JVM_ARGS=-Djava.net.preferIPv4Stack=true KAFKA_OPTS=-Djava.net.preferIPv4Stack=true ZOOKEEPER_OPTS=-Djava.net.preferIPv4Stack=true
HUMIO_JVM_ARGS=-Djava.net.preferIPv6Addresses=true KAFKA_OPTS=-Djava.net.preferIPv6Addresses=true ZOOKEEPER_OPTS=-Djava.net.preferIPv6Addresses=true
In Humio you can detect when a host or some other log source stops
sending logs using the
now() function and
groupby(host, function=max(@timestamp, as=@timestamp)) | missing:=(now()-@timestamp)>(5*60*1000) | missing=true
The above query shows a line for each host that we have not heard from in the last 5 minutes (timestamps in Humio are in milliseconds). You should run the query as a live search in a time interval that is longer than you “missing” threshold - when the last log from a log sources is older than your search time interval the log source will disappear from the result.