Working by patch
The patch is the unit that gets scheduled: one patch closes many vulnerabilities on many servers, and a team patches a machine once.
A team does not schedule a CVE. It schedules a patch: one patch closes several vulnerabilities on several servers, and the machine gets one window rather than a window per vulnerability. That is why By patch is the default mode of the vulnerability register and not a secondary view.
What the row holds
| Column | What it holds |
|---|---|
| Patch | The patch name, with the origin of that name beneath it |
| Vulnerabilities | How many CVEs the patch closes |
| CVE | The identifiers themselves, two and then “+N” |
| Severity | The worst level in the group, with its highest CVSS |
| Servers | How many servers are known to be exposed |
| Systems | Which systems those servers belong to |
| Type | The class labels of the vulnerabilities in the group |
| Product | The affected products |
| Open | How many of the vulnerabilities are still open |
| Published | The latest publication date in the group |
Clicking a row expands it: the vulnerabilities the patch closes, and the servers it has to be applied to.
How a row lands in a group
Grouping may never drop a row, and that rule decides everything else.
- A vulnerability with a recorded patch goes under it. If the feed named several patches for the same vulnerability — different builds of one product — it appears under each of them.
- A vulnerability with no patch but a KB number is grouped by the KB, and the row says so: “From KB / patch”. That is weaker evidence than a patch the feed named, and a row that did not state the difference would be overstating what is known.
- A vulnerability with neither lands in an explicitly named No patch row.
Server names are folded on the first DNS label, so the same machine written once as a short name and once as an FQDN is counted once. A server name the feed gave that the inventory has never heard of is counted separately as Not in inventory and is never dropped — “we do not know what this is” is a finding, not a rounding error.
From the row to a maintenance window
A patch row carries a schedule button, for administrators. It opens the scheduling wizard with that patch's server list already loaded, up to fifty servers — the exact number is written on the button itself. A row with no known servers cannot be scheduled, and the button says that too.
Export
The Excel export of this mode produces two sheets: the patch table, and a second sheet with one row per patch and server — system name, host name, severity and the vulnerabilities. The team doing the work works from the second one. That sheet has a row limit, and if rows did not fit it says how many.
Updated
This page is the file content/docs/en/v1/risk/by-patch.mdx