Guide
Verify a server vulnerability fix
Scan the host, read the evidence, preview a selected update, and verify the result. Keep each step explicit so an update is never confused with a confirmed fix.
Illustrative workflow
Follow the workflow
- 1Inspect packages
- 2Match advisories
- 3Review the update
- 4Verify with a fresh scan
Local consent controls remote access. Unknown coverage stays visible.
Inspect the current server
Run a server scan and review high-severity findings. Check the installed package version, distribution advisory source, coverage, and freshness before deciding which package to update. Application dependency findings are reported separately. If coverage is partial or a source is stale, address that gap before making a clean claim. RPM-based systems use cached repository metadata; refresh it with the distribution tools when required.
Start by deciding which service the finding could affect and who owns the maintenance window. Check the distribution package evidence before choosing an update, because package naming and backported fixes may differ from an upstream advisory. Keep a record of partial coverage so a scan with missing evidence cannot be mistaken for a fully reviewed server.
phelix security server scanphelix security server report --severity highphelix security server preview --package opensslphelix security server apply --preview-token <token>phelix security server verify --operation-id <id>phelix security server reboot-statusReview and apply the selected update
Preview only the packages you intend to change and inspect the package manager result. The preview token lasts 30 minutes and binds the captured package state. Apply the reviewed plan with root privileges. If the package state changes or another package operation holds a lock, apply refuses the operation. Review a new preview rather than forcing the old one.
Read the proposed changes as a maintenance plan, including any related packages the package manager will update. Check that you can recover the host if the update disrupts an essential dependency. The preview binds the observed package state; when that state changes, a fresh review is more useful than trying to reuse an outdated maintenance decision.
Verify and record what remains
Use the operation ID returned by apply to run verification. The fresh scan identifies resolved and remaining findings for each requested package. Review reboot status and schedule any required restart yourself. For remote management, grant reporting consent locally and keep the monitor connected. Read the recorded operation outcome after a reconnect before requesting another update.
After verification, test the application operation supported by the updated package and review whether a restart is still required. Keep resolved and remaining findings together in the maintenance record. A clean package result is evidence about the advisory match, while an application check confirms that the service customers use still works after the maintenance change.
Related
Read next
One binary between git push and production.
Install the CLI, verify the host, and follow the quick start to build your first application.