Introduction
Setting up configure loki for log aggregation on centos stream 10 on a CentOS Stream 10 server is a common task for system administrators, DevOps engineers, and site reliability engineers. This guide explains how to Configure Loki for Log Aggregation on CentOS Stream 10, with all the commands you need, the SELinux and firewalld considerations to keep in mind, and how to validate the result on the live system.
Prerequisites
The procedure assumes you are working on a fully patched CentOS Stream 10 machine. Run sudo dnf upgrade -y before continuing so you start from a known good baseline. SELinux should be in enforcing mode (the default), and firewalld should be running. You will also need a user account with sudo privileges configured in /etc/sudoers.d/.
Step 1: Update CentOS Stream 10 and Enable Repositories
Ensure your CentOS Stream 10 system is fully patched before installing new software. The AppStream repository is enabled by default on registered systems and provides the modular packages needed for most modern workloads. Logging at this stage is critical — anything you do here that is not logged will be very difficult to audit six months from now, so prefer commands that leave a trail in /var/log or the systemd journal.
sudo dnf upgrade -y
sudo dnf repolist enabledStep 2: Install Required Packages
Install the loki package along with any supporting dependencies from the standard CentOS Stream 10 repositories. dnf will automatically resolve and pull in libraries, language runtimes, and configuration files. If you are running this inside a Kickstart automation or an Ansible role, capture every command into version control so the deployment can be re-run from scratch against a fresh VM at any time.
sudo dnf install -y lokiStep 3: Apply the Initial Configuration
Now configure the component for your environment. Always keep a backup copy of the original configuration file so you can roll back quickly if something goes wrong, and prefer editing files in /etc/ over modifying the package defaults inside /usr/share/. Logging at this stage is critical — anything you do here that is not logged will be very difficult to audit six months from now, so prefer commands that leave a trail in /var/log or the systemd journal.
sudo cp /etc/loki/loki.conf /etc/loki/loki.conf.bak
sudo nano /etc/loki/loki.conf # edit settings as requiredStep 4: Enable and Start the loki Service
systemd is responsible for service lifecycle on CentOS Stream 10. The enable --now flag both starts the service immediately and configures it to launch automatically at every boot, which is the behavior you almost always want for a server-side component. Logging at this stage is critical — anything you do here that is not logged will be very difficult to audit six months from now, so prefer commands that leave a trail in /var/log or the systemd journal.
sudo systemctl enable --now loki
sudo systemctl status loki --no-pagerStep 5: Open the Required Firewalld Port
firewalld is the default firewall on CentOS Stream 10 and uses nftables under the hood. Open only the specific port required for this service, and prefer named services over raw port numbers where they exist because they survive port-number changes upstream. Take a moment to read the upstream documentation linked from the package manpage; Red Hat does not always carry every upstream option, and reading the official notes prevents you from copy-pasting flags that silently get ignored.
sudo firewall-cmd --permanent --add-port=3100/tcp
sudo firewall-cmd --reloadStep 6: SELinux Considerations
SELinux runs in enforcing mode by default on CentOS Stream 10. If your service needs to write outside its default directories, bind to non-standard ports, or connect outbound to other services, you will need to set the appropriate boolean or label the files. The commands below are a typical starting point. Logging at this stage is critical — anything you do here that is not logged will be very difficult to audit six months from now, so prefer commands that leave a trail in /var/log or the systemd journal.
sudo semanage port -a -t http_port_t -p tcp 3100
# only if you change the listening port
sudo ausearch -m AVC -ts recentStep 7: Inspect Service Logs
All systemd-managed services on CentOS Stream 10 stream their output to the journal, which is searchable, indexed, and persists across reboots once you create /var/log/journal. Use journalctl to follow logs in real time and to investigate startup failures. If you are running this inside a Kickstart automation or an Ansible role, capture every command into version control so the deployment can be re-run from scratch against a fresh VM at any time.
sudo journalctl -u loki -e --no-pager
sudo journalctl -u loki -fTroubleshooting Common Issues
If the service refuses to start, the first place to look is the systemd journal — every service on CentOS Stream 10 logs there by default. Filter to the last boot to avoid wading through historical entries. The second most common class of problem on a fresh install is SELinux denials, especially when a service tries to read from or write to a directory that is not labelled with its expected type. Use ausearch -m AVC -ts recent to look for denials, and either set the correct file context with semanage fcontext + restorecon or flip the relevant boolean. Finally, if the service starts but is unreachable, double-check firewalld with firewall-cmd --list-all and confirm the runtime configuration matches the permanent one.
sudo journalctl -b --priority=err
sudo ausearch -m AVC -ts recent
sudo firewall-cmd --list-allBest Practices and Hardening
For any production deployment on CentOS Stream 10 you should track configuration in a version control system, apply security errata regularly with dnf-automatic, and centralise log collection so that a compromised host cannot quietly erase its own audit trail. Run periodic OpenSCAP compliance scans against the CIS or DISA STIG profile to catch drift. If the service exposes a network port, place it behind a reverse proxy or VPN where possible and rotate any credentials it uses on a schedule. Snapshot the system (using Stratis, LVM, or your hypervisor) before every major change so you have a fast rollback path.
sudo dnf install -y dnf-automatic
sudo systemctl enable --now dnf-automatic.timer
sudo oscap xccdf eval --profile xccdf_org.ssgproject.content_profile_cis /usr/share/xml/scap/ssg/content/ssg-centos_stream10-ds.xmlVerification
After completing every step, run a quick set of checks to confirm the deployment is healthy on CentOS Stream 10. Examine the systemd unit state to make sure no units have failed, look for any SELinux denials in the audit log, inspect the listening sockets to confirm the service is bound to the expected interface and port, and finally make a real client request to validate end-to-end functionality. If any of those four checks fail, return to the troubleshooting section before treating the deployment as complete.
sudo systemctl --failed
sudo ausearch -m AVC -ts recent || true
sudo ss -tulpn
sudo journalctl --since "10 minutes ago" --priority=warningConclusion
That completes the walk-through of how to Configure Loki for Log Aggregation on CentOS Stream 10 on CentOS Stream 10. The end result is a reproducible deployment that you can encode in Ansible or Tekton pipelines, and that benefits from the long-term support and security errata Red Hat provides. Consider exporting your final configuration files to a Git repository so you can version them.
Looking forward, consider encoding the steps above as an Ansible role so the procedure becomes reproducible across your entire fleet, and add a Prometheus scrape config (or a Zabbix template) so the service is monitored from the moment it starts. Pair the deployment with a backup strategy — restic, borgbackup, or rsnapshot all work well on CentOS Stream 10 — so that recovery from data loss is a matter of minutes rather than hours.
